DIY Micro AM RX

[clipclop]

I have been working on shrinking the old Albin circuite useing modern SMT components to make a micro RX

[clipclop]

Heres the circuite i have running , range is over 40 meters quite good enough for indoor use

[clipclop]

Heres my prototype PCB , 604 sized SMT caps and resistors , sot-23 transistors BC848 , National LMV301 SC70 package op_amp
PCB pattern removed

Stewart

[KreAture]

Hmm, here's my FM receiver module:
http://www.unixcore.com/~kreature/images/rex/rex_rx_ok2.jpg

It does not decode the individual channels to seperate pinns nor does it amplify it yet. The image only shows a quick beta build but it measures less than 17x17mm and weight is less than 0.8grams. I am planning to make it include a small ESC as well as some multi-purpose drivers for both actuators and micro servos.

Currently I am having difficulty getting the selectivity / channel seperation good enough. Sensitivity is around 6uV btw.

[clipclop]

KreAture,
I am aiming to get this RX to about 10mm X 10mm and less than 500mg useing 403 SMT C's & R's and mini SMT transistors 1/2 size of sot-23 , the op-amp is SC70 size which is 1/2 sot-23 size the RX will have the PIC ant esc fet on the back .
Stewart

[KreAture]

NICE clipclop!

Almost makes me wish I had a AM tx LOL.

[johnnie7]

lookin good
guess ill need get me soldering out soon:D

[JRuggiero]

clipclop and KreAture,

You've both done some intriguing work. How about looking into hacking a BitCharger or similar receiver so it'll work on frequencies other than 49, 35, or 27 megaHertz? Add more channels/functions? Increase selectivity?

Jim R

[clipclop]

Heres the parts placement ,

Its easier to build a new RX than try and modify the bit RX , which is basically the same around the first transistor the chip on the bit RX is the problem.
NOTE!!!! there is an error in the op-amp pinout connections
a new PCB is in the works
The cct does work , I just goofed on the PCB

[rRoberto Grassi]

Very interesting!!, what I couldn't see is the working freq.; and what about the PIC programming?

[clipclop]

rRoberto,
It will tune 27-30mhz , the pic programe is from Koichi Tanaka's DIY IR thread here http://www.rcgroups.com/forums/showthread.php?threadid=24711&goto=newpost
Stewart

[rRoberto Grassi]

Thanks, Stewart; as I'm quite poor in RF, for changing it to 40 or at least 72Mhz, what to do?

[clipclop]

rRoberto,
to raise the frequency of operation you need to decrease the no of turns on L1 , I havent posted coil details yet , I will and I'll post theretical coil specks for 40 and 72 megs as well .
40 megs should work ok as there is little difference in coil details but 72 may need some of the other cap values adjusting to balance the circuite and maintain oscilation .
I just did a rough calculation for 40 megs with my coil if the C2 cap was changed to 10pf it should work ok .
Stewart

[rRoberto Grassi]

Thanks clipclop, I'll wait for the 72mhz coil and cap, 'cause forgot that my only TX in AM is that one - you know,at 52 years old my memory starts making bad jokes-

[clipclop]

L1 coil specks
T he former used has outside dia 4.83mm height 14mm with a 4mmx10mm ferrite slug 12 1/2 turns of .5mm copper wire

Have a look in this site there is a calculator that works ofline if you save the page ,it calculates caps ,coils frequency its very handy with this you should be able to work out the specks you need :- http://w1.859.telia.com/~u85920178/blocks/osc7m00.htm

Another Albin used this coil

27 mhz coil is wound on a paper tube.
The dimensions of the ferrite slug are: 3.76 mm diameter x 5.59 mm
length.
The coil is 11 1/2 turns of .29 mm wire.

[clipclop]

Heres a new PCB layout .
Note this RX uses a national LM7301 op-amp instead of the LMV301 the pinouts are different . a mask will follow and an updated parts list .
Also note this PCB layout uses SMT inductors and a trimpot to tune .
This PCB layout was CAD drawn for us by another member of the forum.
The PCB is aprox 13mm X 22mm

[clipclop]

Here is the PCB mask , this and the previous CAD drawing was kindly done by fellow forum member Dave Burley . I would like to take this moment to thank Dave for his great work in doing these drawings for us .
Note the 3 805 sized capacitors are mounted on there sides to save bourd space , and sujested transistors are BC847B or C , or BC848B or C types also due to differences in transistor hfe its sujested when building to fit a trimpot in place of the 8k resistor set it for best operation then fit the closest resistor . I have found values from 4.2k to 8.6k were needed or alternatively try fitting a 4.7k -5.6k and try it .
The cct works at 27megs and 29megs other frequencies have not been tried yet .
Stewart

[clipclop]

Finnal version update .
The PCB layout has been changed to include a esc output for motor and the access to the actuator outputs has been improved . If this RX is built on thin PCB it should finnish up around the gram quite good enough for micro flight .
These CAD drawings were kindly done by dburley , Thanks Dave
Have fun and DIY
Stewart

[clipclop]

PCB mask.
Note:- For those interested in 72m if a 1.5uH inductor and 1.5-3pF trimcap are used it theretically should tune 72megs

[clipclop]

Final mod:- a 10pf capacitor was added in parrallel to to L1 and the trimcap , this will reduce the tuneable range but make tuneing easier . this mod is for 27-29megs

[clipclop]

Modded mask

[KreAture]

How large is that layout now clipclop ?
My FM rx was only 17x17mm btw.

I should get back in the saddle on it and try to make it decode the channels. That would allow me to test it with some servos attached.

[clipclop]

The PCB is 12.62mm X 22.16mm if .3mm PCB is used the RX should be about 1gram .
Stewart

[clipclop]

Kreature,
I would sugest an op-amp setup similar to mine to interface between the RX chip and the pic the SC70 sized LMV301 is small and has good signal responce and gain . the inverting config gives better signal from small inputs so your decode needs to be for inverted signal , assumeing non-inverted tx signal .
Stewart

[KreAture]

Yep. I will be using a opamp of some sorts. I may use a smaller one though. They come in SOT-223 you know.

[clipclop]

Kreature,
SC70 is 1/2 size of sot23 how big is a sot223?

[KreAture]

Ahh, sorry I was confusing the PIC decoder with the opamp. I see it now :)

[JMP_blackfoot]

Originally posted by clipclop
PCB mask.
Note:- For those interested in 72m if a 1.5uH inductor and 1.5-3pF trimcap are used it theretically should tune 72megs
If I were building one, I'd use an inductor of lower value, around 0.2~0.5 uH.
To change from 27 to 72 MHz, I would tend to divide the coil and capacitor each by the frequency ratio, i.e. 72/27 = 2.67 -> divide coil and capacitor by 2.67. The idea is to keep the impedance ratios similar, if you start from a successful design.
My 2 (Euro) cents worth :cool:

[clipclop]

JMP ,
Thanks you are prolly right , I havent tested on 72megs , its not a frequency used here in Australia so I have no TX to test that Frequency .
Stewart

[clipclop]

Just Etched a PCB from .3mm doublesided board , its .2mm and under 100mm-the min my digital scales will read , it reads 0.0 after settling .
So the RX should be under 500mg , pretty light .
Stewart

[vaughan@world.s]

Stewart,

Could you tell me your source vendor for the paper tube/ferrite slugs for your L1 tuning coil and your source for .3mm doublesided board. This stuff is hard to find.

Thanks

Brian

[kit]

.026 inch double sided pc board, cheap.

http://sales.goldmine-elec.com/prodlist.asp?catid=2218

[clipclop]

Brian,
paper tube is hand made to suit the dia of the ferrite slug , its a 3mm dia F29 ferrite . I got mine from a local electronics store here in australia called Dick Smiths .
In my latest PCB layout I am useing SMT chokes from digikey with a SMT trimcap for tuneing .
Stewart

[vaughan@world.s]

Thanks kit and Stewart. I already have the pcb material on order.

I can remember as a kid how easy it was to get coil forms and variable caps. With the introduction of PLL radios it has all gone away.

Brian

[clipclop]

Hi!! all,
Just had my RX working made up on the last PCB layout .
I did need to change 1 component value from the breadboarded cct as it wouldnt go into regen osscillation , I thought maybe C3 may need changeing but it is fine at 22pF .
C1 needs to be 22pF not 33pF as shown in the cct .
I tested both BC848B and BC847B transistors either will work fine , although I havent tested BC849B it should also work as its a closer match to the BC848 than the BC847 and it works .

The inductors I used in the RX are SMT types .
L1 is digikey 445-1016-1 1.8uH
L2 " " 445-1029-1 22uH
C1 is 22pF
C2 is digikey 490-2003-1 SMT trimcap 4.5-20pF
It weighs 300mg without the Pic installed so completed RX should be about 400mg
Stewart

[clipclop]

Updated CCT with finnal component values as used on the last PCB layout .
I've built this and it works ,
Updated today 18-04-04 added C11 to cct . Note it is on the PCB I just missed updateing the cct
Stewart
cct removed see latest

[MAVA]

Stewart,

Thank you for you have done for this thread,

Picked up alot of Ideas from your postings.

I do have a simple question, are you still getting 40meters or more of range?

Martin

PS.. I am in the USA so I will need to modify it for 72MHZ, I have an idea how to do it

[clipclop]

Martin,
you will need to decrease the inductance of L1 as well as use a lower value trimcap .
unfortunately I havent a 72 meg Tx -we dont use that band here in Australia .
you may find that after getting the RC cct tuned you may need to alter C2 and C3 in order to get regeneration .
Has anyone built an albin based RX on 72 megs ???
The range is still over 20 meters with the TX antenna down .

[vintage1]

Roughlty halve the value of all L's and C's round the front end to get up to 72Mhz.

Be aware these recivers are NOT good at flying in company, not only do they pick up the strongest transmitter around, they also radiate enormous amounts of trash across the band when they are not 'locked on'.

In my youth I built several superegens, and solved the problem by building the strongest transmitter around (by an illegal factor of about ten :D)

[clipclop]

yup they for solo flying only , but who wants there super micro flying wiff UFO's

[clipclop]

Almost finnished a second RX today , its on 0.2mm PCB weighs just under 300mg only the solder for the pic and a litz wire antenna to be added , it was weighed with the pic . This one is lighter due to thinner PCB and neater soldering .
Stewart

[MAVA]

ClipClop,

Got Pictures of your new RX ?

Dumb question...What Is "Litz Wire"? It is the type of wire used to wire magnetic actuators to control surfaces?

Martin

[AndyOne]

Martin,

Litz wire is a stranded kind where each individual strand is insulated, it reduces the "skin effect" (works better at high frequency). It's not necessary for connecting DC devices like actuators other than for reducing fatigue breaks.

Vintage1,
Wouldn't it be best to keep L2 at 22uH as I believe this inductor controls the regeneration frequency and not the received frequency?

Andy.

[clipclop]

Vintage1,
Andy is correct only L1, C2 and the new 10pF cap need changeing to alter the band .
Stewart

[clipclop]

Heres the Parts List with DIGIKEY part No's .
I havent include the Pic I used a PIC12C509A-04/SN loaded with Koichi's Futaba IRX decode
Cheers and have fun
Added Pic and Fet , also the R across the fet to the parts list
27-4-04 removed parts list, see latest
Stewart

[vintage1]

Apolologies...for an over simplistic answer. but I think C1 C3 and C4 would also need changing. The RFC is I agree not part of the RF side of things..when I used to build these things we made C3 out of two bits of wire twisted togeher sometimes...and cut shorter and shorter till it just stopped working, then twisted it up a bit tighter :D

Use of too large an aerial coupling capacitor (C4) makes these things very sensitive to things like hands around around near them. Of course we had a different set of conditions - sometimes we were just using pulsed carrier and when the carrier was off, they would pick up interference...tone use or digital use is better.

C1 is an odd one. I never worked out what it did. But hey,if you took it out it stopped working :D SoI am not sure whether it would need changing for 72Mhz use.

[clipclop]

Vintage1,
C1 is tied to L2,C5,D1 and the transistor C3 is transistor dependent and not really frequency dependent .
I have tried values of C4 between 5-10pF with no noticeable difference . The RX seems resonably free of the effects of handling when tuned onto frequency .
Even a metal tuneing tool seems to have minimal effect .Although it does pick up the local radio station load and clear if you hold a metal probe and touch it to the base of T1 -my son was most amazed lol
Its a pity I havent a 72meg Tx I'de set it up for that frequency too .
Hope you dont have too manny probs setting it to 72megs .
Cheers
Stewart

[clipclop]

Finnal corrections the Rfet should be 20k not 30k heres the corrected cct and parts list . 30k should still work ok its to stop the motor fet turning on dureing pic powerup .
26th corrected C1-3 digikey No

[maulionc]

helloo clipclop

have you tested your RX design ??? is really available???is designed for AM and FM , FM or AM???

thanks :rolleyes:

[clipclop]

Maulionc,
Its AM but "may" work with FM , I havent tried tuneing my FM TX to it yet .
I have driven servos and a ESC driveing a speed 280 motor , the RX worked good . it was powered by the ESC Bec at that time .
Stewart

[tru168]

Hi guys,The receiver looks quite interesting,i will build one for my indoor plane.

FYI,the superegen type receiver can receive AM and FM without any modification.I had build several (not SMD type, larger version) few years ago for my parkflyers and using SANWA FM transmitter.

[clipclop]

New updates L1 changed to 1.5uh , C2 changed to 2.5-6pF trimmer , Cx changed to 15pF for 29m band 18pF for 27m band . Revoved files -see latest

[clipclop]

JoeP has scoped the signals in the RX here they are:-
Pic1 is baseT2
Pic2 is collector T2
Pic3 Is input to PIC
Note the above modifications - D2 removeal C6 .47mF were made as a result of these scope tests .
L1, C2 Cx mods were to make tuneing easier
Stewart

[clipclop]

picture of the RX dueing testing , not a good picture but you can see the size at least .
Stewart

[Bruce Abbott]

The input signal to the PIC is showing some oscillation and slew rate limiting on downward transitions, probabably due to excessive capacitive loading. Try reducing the value of C10 and/or adding a resistor between the output of U1 and C10.

[clipclop]

Bruce,
I did try a 1nF cap for C10 , but that was befor adding the 10uF across the rails .
It may be a good idear for me to retest the 1nF .
The LM7301 op-amp is supposed to be able to handle high capacitive loads esspescialy at high gains .
Without haveing a scope there is only so much I can acheave listening to the signal .
If anyone with a scope would like to breadboard the RX and try diff values for C9,C10 ,and maybe test a cap across the base emitter of T2 Please do .
Overall I am happy with the signal its within the Pic's input requirements , but if improvements can be made without adding parts that would be good to .
Stewart

[Joe P]

Bruce,
What looks like oscillation on the downward transition are cause be the superregen
quench frequency leaking through, if you look at the base of t2 it's quite strong
and hard to filter out.
I don't think that it's a problem for the PIC. But i'm having a problem with throttle
throw I'm using Koichi's hex code the was for his GWS to actuator conversion
an i'm using an old Futaba FG radio with 1.2 msec pulse widths the actuator ports
are ok can go full on to off on elevator and aileron PIC outputs.

[clipclop]

JoeP,
Have a look at Darkiths code to , it may be of use :- http://darkith.dyndns.org/~darkith/html/ir.shtml
I am in the throws of learning a bit of programming , hopefully a more universal code will be the result . My old AM JR TX is a problem as well I can get it driveing servos , but the actuator code is not suitable .
I think we need a more intuative code to detect some of the TX settings as the old AM TX's varried some what.
hopefully it wont be long to code something .
Stewart

[clipclop]

Some new modifications to improve the signal , all will fit the PCB .
R5 3.3k ,
Removeall D2, C8
The Rx worked with the previous values these improve the signal .
Stewart
27th May , removed some past cct's and updated the parts list.
NOTE!!!
Some of the changes have been revised , Do not fitt RC7 @ 3.3k , it KILLS the signal
Stewart

[KreAture]

clipclop, get that pc-oscilloscope program for soundcard...
It will give a samplerate of 48 kHz I think, and should be ok for the PPM signal.
(It won't do for RF signal, but for input to PIC it will be ok and better than nothing.)

Use MIC input to load the circuit little. Might even use a 1 mOhm series resistor and let the internal gain of the soundcard do some amping. Should allow you to test those cap values.

[clipclop]

Kreature,
I tried 2 different soundcard programs :-( my sound card must be capacitive coupled the wave form was triangular in both programs no matter what settings .
but thanks anyway , I need to find a cheap sound card apparently the scope programs work with most of those .
Stewart

[clipclop]

Parts list and CCT have been updated to the new component values ,
Heres a parts placement Drg to aid construction .
Stewart

[clipclop]

Please note the previous changes to C9 and the removal of C10 have been reversed , joep found that the signal wasnt stable when driveing the pic so he has changed his recomendation on these .
He also found that removeall of C8 gives a better signal as the output is squelched by RC7 when there is no signal , it also allows RC7 to pull the lows to the -Ve giveing more possitive switching to the Pic.
Also revised parts list attached
Stewart
Note RC7@3.3k KILLS the signal -do not fit !
I will further test RC7 to find a suitable value or if its needed at all .
Also for 27-36megs C6 is best at 10nf

[clipclop]

I must appologise for some of the previous changes , they were recomended to me and I posted them without testing them myself .
The fitting of RC7@4.7k will kill the signal the value is way too low .
The RX works without it , I will test to see what value is suitable or if its needed at all .
Re C6 for 27-36megs 10nF is best not .47uF
R5 is good at 3.3k
C8 , RX works with or without it , I will scope test to see whats best .
I am in the process of building 3 Rx's for scope testing at my local high school .
The aim of the tests are :-
1/ to fine tune the RX values
2/ to test the RX at 36megs for FM reception .
3/ to get the signal of my AM TX so suitable code can be written for the Pic

Update:
New cct with current values for SC70 &sot23 transistors types .
also note added re CX value - due to the nature of the regenerative cct, parts placement and PCB track size variations will induce different inductnce and capacitance to the cct that may require a different value for CX in order to tune the required frequency .
Not noted on cct , in order to get the RX tuned easier the C2 variable cap may be changed to 3-10pF with a suitable adjustment to CX will give a wider range of tuneing . It seems that some people are not getting the RX on frequency due to variations in PCB 's and layout .
or the C2 can be 8-25pF and leave out CX altogether adjustment of C2 is a little touchy but I have built and tuned the RX with this varicap.

[clipclop]

note updates above ,also see developement in this thread http://www.rcgroups.com/forums/showthread.php?t=189805&goto=newpost
Stewart