RF remote control circuit

Hi Zsolt,
I cannot say what the function of each coil is without a schematic. I guess by their size, that the air coil may be the oscillator one and the green coil an IF filter. It may be another way as well. In my previous advice, I had mainly the air coil in mind. I cannot judge how stable the green coil is without seeing how it is made. It may also require reconstruction, maybe to a fixed value (and adding a small capacitive trimmer for tuning).

The air wound coil changes its size and hence its inductance with temperature considerably. The goal is to stop these changes. If some tension builds in the wire with temperature changes, but the size stays constant, you are O.K.

The material used to reinforce the coil(s) must be strong - the wire tends to move with great force. It may not degrade the coil Q. Unfilled epoxy perhaps may do.

Small value SMD caps are ceramic. The key is the temperature coefficient (TC) of the dielectric inside. The TC may be negative or positive, small or larger in absolute value, depending on the ceramic mix. Coils have always TC positive. You need capacitors with negative TC to compensate the drift - but do not overcompensate.

VBR from Ivan

hello ,
i will try out the pointed issues from below after i get the schematic . Is the adjusting on the green coil somehow compensate the air wound coil degradation ?
Never touched the air wound coil. I saw in some older radios coils like that with a kind of sponge inside being all impregnated with some sort of hard wax . I do have some bicomponent glue that hardens like concrete . Ceramic capacitors are better than those smd caps on the board ?

Hello Zsolt,
As Ivan pointed out, air-spaced coils can be very sensitive to changes and putting them on a former can dramatically reduce mechanical changes, and microphony.

One suggestion, when I made the FM wireless microphone I had similar problems, and people assembling the kits could never wind the same coil twice. You could also shout at the board without a microphone and hear the recovered audio quite nicely.

As for frequency stability with metal expansion, I tried using iron-based wire (the same as you get on many resistors), since the expansion coefficient is half that of copper. It improved frequency drift a lot but did not eliminate it.

Then I tried etching the coil on a bit of PCB. What a difference that made. Every coil was exactly identical, cannot be deformed (and de-tuned) with pressure, and although copper, it still showed very little frequency change with temperature. Perhaps the copper was forced to expand sideways? Whatever, it worked for both repeatability and stability.



Another point, the "gimmik" capacitor can have a small negative temperature coefficient, depending what type of insulator you use. In the picture I used telephone IDM interconnecting wire. PVC hook-up wire had a positive coefficient.

Hope this suggestion helps.

Very best regards - Harry

Hi,
maybe yes, maybe no. It is necessary to trace the schematics or find it somewhere and play with those capacitors which affect the working frequency (probably 1 or 2 pieces only).
I see the coil is airwound, its turns can move freely with temperature changes. I would start by rewinding it very tightly on a ceramic former (no plastic) and fixing it so that it keeps its size and shape. Some authors recommend to heat the wire before and during winding and let it shrink on the finished coil. This alone may help.

VBR from Ivan

hello 
My thinking was to hat up the board to more then 20 , lets say 30 C and make it work like that. In the summer it will be hot anyway so the thermostatic circuit will switch off by itself and the board will keep working since it's hot.
Will changing those smd caps make a change ?
This is the little rx board 
https://www.banggood.com/Geekcreit-433Mhz-RF-Decoder-Transmitter-With-Receiver-Module-Kit-For-ARM-MCU-Wireless-Geekcreit-for-Arduino-products-that-work-with-official-Arduino-boards-p-74102.html?cur_warehouse=CN

Hi Zsolt,
thermostating the receiver is possible, but it would increase the power consumption considerably. Keeping the receiver chamber on some 20°C and heating in in winter only is slightly better, as well as cooling it in hot summer days using a Peltier element. The power-saving way is to compensate the temperature drift by using ceramic capacitors with negative temperature coefficient (NPO) where appropriate*. This would require days of fiddling with the circuit, fridge and a soldering iron, but should make the circuit work in a wider temperature range.

*those capacitors, which are a part of the resonant circuit and affect the working frequency

VBR from Ivan

hello to all
Im back with my rf control (from2019) circuit again... It is working, just that i have some annoying   frequency drift on the rx side i think. The tx is still mounted on the car (and it broke my car radio )and thanks to the field strength meter i can tell it is working ok.
I always have to adjust the rx coil, once in spring and once in winter like now. It is a really tedious operation due to distances and climbing.
I thought about setting up the little receiver board in a thermostatic chamber. Like getting it in a little sealed  aluminium box with a heater resistor and put in there a simple temperature control circuit set up to about  30 'C or so. I believe it will keep it steady in the winter and it will be warm in the summer anyway.
Any opinions about this?

zsolt wrote:Hi,
again fog is starting in the mornings because of the lake. I believe the fog affects my radio link between car and gate. The gate starts operating at second command sometimes.
I wonder if it is water getting to the rx circuit or simply it's because of the water in the air.

Hi,
foggy weather may cause:
- water condensing in the receiver, lowering its sensitivity or rendering it completely inoperational;
- water condensing on insulators of the antenna, changing its parameters;
- wet surfaces near the line between the TX and RX create reflections and multiple paths for the RF signal, causing interference;
- fog in the air increases the losses of RF waves, but IMHO this effect is small below 1 GHz.

BR from Ivan

Hi,
again fog is starting in the mornings because of the lake. I believe the fog affects my radio link between car and gate. The gate starts operating at second command sometimes.
I wonder if it is water getting to the rx circuit or simply it's because of the water in the air.

You can feed the amplified signal to the same dipole antenna. Range should increase 4X or 1/2 Km.

ok. i give it a try when i get my tranzistor. My tx antenna is a 433 mhz dipole. That can not be changed because it's quite perfectly installed (looks like from factory). From my lately  experiments the antenna makes a lot of difference and it seems to be neglected completely in documentations of the little modules.

PDF of my article. Just copy the amplifier section.
RFC is a few turns on a small binocular ferrite core. They tested the circuit for only 100m but it works well up to half a kms away or 400-500m. I used the lead of a resistor for the 1.5cm wire link in the output matching section. Build it ugly bug style on a copper clad board with a large ground and it will work the first time. Output power = 0.3W at 433MHz

If BFG135 is hard to find then BFG591 can be used instead. They are cheap on eBay from China.

I am thinking about tweaking the amplifier again for more power output. Will experiment when I get time.

https://docdro.id/XPXJPIf

hi,
of course i want circuit  

I used a BFG135 transistor to boost the output of 433MHz model to 300mW and the range was 400 meters with a 17cm wire antenna. You want the circuit? It's a simple one transistor amplifier.

Hi,
i suspected that the antenna gain word is not used in the same sense as amplifier gain (which has power supply available to get energy from)
One in all i get better and better in 433Mhz with my remote. Now it doesn't matter if it rains, it's fog or is cold. The gate operates every time. Excluding the coax between the simple dipole and the small tx board increased the range a lot.
It remains to correct the rx side of the remote control and it will be great.
I made an other unit to experiment the rx part of the remote control (the one painted red) untill i get it right, so i don't need to climb the roof to make adjustments than go back to the car ( 'bentley'   ) 
I saw a lot of pictures with shorted stubs baluns and loops and things. My experiment rx antenna is 1/2 dipole and the part uniting them 1/4 +1/4. That's how i saw the picture. But i think there might be error. Like in the picture i cut the dipole so that it's extremities are 1/2, not the elements individually make the 1/2 length. It s going to take a while but i get my 100m. 
If not i put the rx on my neighbours first fence and than put a 40m cable to my gate. Before this radical solution i want to make the most out of these little modules.
Other modules that promise more range have quite elaborated antennas and i think that's the secret beside the tx power.
Anyway no commercial gate opener has the range i already have so my neighbours (driving more expensive 'bentleys') have a thing to envy 

zsolt wrote: I taught only tranzistors and valves have gain...

Antennas do not have a gain, in that you get out something for nothing.
An "Isotropic" (theoretical antenna - does not exist) radiates a signal uniformly in all directions and has a loss of about 2.4dBd (just under 1/2) compared to a 1/2-wave dipole.

The dipole does not radiate power along it's length, that means the radiation around it at 90' to is almost twice as strong.

A 6-element Yagi antenna pushes nearly ALL the signal in only one direction, so in that direction the signal is more concentrated. You still get the same power out, but all in one direction. It is a bit like putting your finger over the end of a hose-pipe - same amount of water, but the poor guy in front of you gets a lot more of it. The guys to the side don't get any. In the case of the 6-element Yagi, there is only 10% of the power leaking to the sides and back, the other 90% is all in one direction, so it appears to give more power in one direction COMPARED TO THE ISOTROPIC antenna that spreads all over the place = 10dbi "gain", or 7.6dBd gain (compared to a dipole).

Best regards from Harry - SM0VPO

Ivan wrote:The lengths of the elements make the difference:
1/ The length of each leg of a dipole is 1/4 of wavelength. The length of each leg of a collinear is 1/2 of wavelength.
Consequently:
2/ The impedance in the center of a dipole is cca 72 ohm balanced. The impedance in the center of a collinear is very high (infinity), balanced.
3/ A dipole requires 1:1 balun for matching to a coax. The balun can be made in different variants, e.g. as a loop of coax, as piece of Lecher conduit etc.
A collinear requires an impedance transformer "infinity to 50 ohms" plus a 1:1 balun. The transformer can be formed as a shorted stub of Lecher conduit, the output of which is exactly at the point where the necessary impedance is - not "somewhwere".
4/ A dipole has logically no gain against a dipole, 0dBd.
A collinear has gain +3,5 dBd against a dipole.

Do you still think that a dipole and a collinear are the same thing?

BR from Ivan

 this is going to take a while. I taught only tranzistors and valves have gain... 
meantime i changed only the tx. I moved the little board right on the dipole without coax. Still on the back of the rear view mirror. Now the distance increased to 2 public lighting pillars. I don't know how much is that in meters but if i double that it's enough
The rx is the same i did not try out the new red one yet.

The lengths of the elements make the difference:
1/ The length of each leg of a dipole is 1/4 of wavelength. The length of each leg of a collinear is 1/2 of wavelength.
Consequently:
2/ The impedance in the center of a dipole is cca 72 ohm balanced. The impedance in the center of a collinear is very high (infinity), balanced.
3/ A dipole requires 1:1 balun for matching to a coax. The balun can be made in different variants, e.g. as a loop of coax, as piece of Lecher conduit etc.
A collinear requires an impedance transformer "infinity to 50 ohms" plus a 1:1 balun. The transformer can be formed as a shorted stub of Lecher conduit, the output of which is exactly at the point where the necessary impedance is - not "somewhwere".
4/ A dipole has logically no gain against a dipole, 0dBd.
A collinear has gain +3,5 dBd against a dipole.

Do you still think that a dipole and a collinear are the same thing?

BR from Ivan

hi, 
well... i see both pictures figuring the same thing. Of my point of view  it's a T shape made of cooper pipe. One connects his wires some where  on the part called stub, an other puts his wires on the part not called stub.... and there is the little coax loop which  i knew from tv antennas.
Is the first time i consider the antenna a part of something so i dont understand a lot. Actually high frequency electromagnetic  phenomenons where never studied by me.All i know in ac is related to 50hz electric power systems. 
That antenna seemed indeed very robust and it was easy to make. When it's time i give it a try as it is. Than if i see that it could be more i make the changes

Hi,
that antenna is a classical dipole with a quarterwave balun, nothing more. The balun is made of tubing instead of coax to obtain mechanical rigidity. You abandoned the idea of a collinear antenna with gain - why?
Adding 1 - 2 turns coupled with L1 (like a transformer) will give you a balanced input of the Rx. But having an antenna with balun, you do not need it.

BR from Ivan

Ivan wrote:

zsolt wrote:The 4 segments have all most same length. Except the little one in the middle.

Now I mentioned this. What is their length? Is it half of wavelength, quarter of wavelength or something else?

hi,
it's exactly the first image, calculated  for 433MHz. I have access to the coil. It is 4 or 5 turns air wound about 2 mm diameter.

zsolt wrote:The 4 segments have all most same length. Except the little one in the middle.

Now I mentioned this. What is their length? Is it half of wavelength, quarter of wavelength or something else?

Hi,
it even may work, but it probably has has poor sensitivity.
ad 1 - Looking at the schematic, the Rx has an unbalanced antenna input. If you have access to L1, it would be the easiest solution to make inductive coupling to that. Both ends of the turn added would be a balanced input.
ad 2 - The active elements are half wavelength each? Then you have the Rx in the high impedance point, unmatched, getting very little signal from the antenna. The stub does not match it. To get it work, the Rx (modified for balanced input, see 1) must be moved along the stub to the point, where its input impedance matches that presented on the stub. This will enable maximum signal transfer into the Rx.
The impedance on a quarter wave shorted stub loaded wth resistance R varies from R to zero according to the sine law.

BR from Ivan

Ivan wrote:The antenna is pretty. Congrats!
1/ The RX has a balanced input? The stub is balanced line, therefore it is wrong to connect any unbalanced load to it.
2/ How did you find the point on the stub, where both impedances match each other?

BR from Ivan

well.... i have my secrets 
1/ balanced input? don't know
2/impedance match? don't know
I just googled and found the antenna. Hope it works 
The 4 segments have all most same length. Except the little one in the middle. That is 4 cm. 
The rx is not on the stub. Is on the edge where the stub begins and dipoles end. That's how i saw a picture online.
I think it was this


i think this is the rx

The antenna is pretty. Congrats!
1/ The RX has a balanced input? The stub is balanced line, therefore it is wrong to connect any unbalanced load to it.
2/ How did you find the point on the stub, where both impedances match each other?

BR from Ivan

new rx right on the antenna . No coax . 

hi 
the meter is not important for me now. The shipment costs more anyway and i would use it just once.
With the rf meter i found out that my tx works.
Meantime i discovered that the system is sensible to temperature not water.
First i taught its the water. In the morning not working, afternoon   working. Today afternoon we have heavy rain and it worked. So it's not the water. It's the temperature. 
If i manage to make the mentioned rx change i will put a few watt resistor in the case of the radio to keep it warm. Then see what happens.
I just had time to cut in angle the cooper pipe and solder the cuts/parts The antenna is ready and painted red

About the EMF meter: looking through the reactions, a buyer mentioned that he has compared the advertized meter with a professional one, and that the specification should be: 5 Hz - 3,5 MHz (So not GHz!)  It seems (according to his remark) that the meter isn't indicating very much above 3500 Kz. Just sayin'...

hi
i made a new antenna  from cooper pipe. Dipole and stub and everything. Exactly 1/2 and 1/4 parts measured with caliper..Since the rx is so small i will put it on the dipole so no coax is needed. If i manage to isolate the thing properly with silicone maybe or some resin i will have only 3 wire shielded cable coming to the microcontroller.

Hi,
I found this EMF meter . It should work wideband up to 3,5 GHz. The price is reasonable. Maybe it can help evaluating antennas.

VBR from Ivan

Hi Ivan,
I have not read through the whole of this thread, but your comments:

"You cut a bar half wavelength plus some cm long. You bend it into U-shape. This way you get a quarter wavelenth shorted stub, used as a phasing element."

You are 100% correct as I understand it. This is what I did with my 6db VHF antenna : http://sm0vpo.altervista.org/antennas/6dbvhf0.htm
It is a 1/2-wave over 1/2-wave and uses a 1/2-wave, folded as a shorted stub, to give a HiZ feed to the ends of the two 1/2-wave elements.

Very best regards fron Harry

zsolt wrote:First statement : "Each element and the stub are all a calculated 1/2-wave in length"  
after a few rows below a second statement : "The only important bit is that the stub should be 1/4-wave in length, not counting the shorted (folded) end." 
Now is it 1/2 or 1/4?

Hi,
this is a question for Harry.
My opinion is: you cut a bar half wavelength plus some cm long. You bend it into U-shape. This way you get a quarter wavelenth shorted stub, used as a phasing element. Two other bars cut exactly to half wavelength long each are not bent and serve as end fed radiators.

VBR from Ivan

i used all possible wifi modules with arduino. I am using 1 wifi repeater to cover 50 m , Without that the wifi itself covers 25...30. 
The same rumors are about the 433mhz  rf modules also, but practice shows it's hard to get 100m. Thous at my first test' s i had 100m with the tx in my hand and a battery
Now i will focus on rx antenna. Maybe after my morning coffee i will search for cooper pipes.

zsolt wrote:i will try that smovpo antenna at the rx next week, but the tx on the bentley  remains like it is because it's best positioning now. The rf meter tells me that is working. Every time i operate the tx it lights up even if the gate opens or not so only range issues can be found there.
Using internet switch is last option because i don't want to relay on gsm network.
I have a measuring device wich i used to position a parabolic dish. It is scaled in db and also beeps. Can i use that somehow to measure if i have the 6 db?

Hi Zsolt,
it is valuable to know that the TX works properly and radiates some RF power. It would give sense to check the raw receiver output now. Is the data corrupt or there is no answer at all?
A satellite dish positioning device should be a simple RF voltmeter in the range 950 - 2150 MHz, so it may indicate the power on 433 MHz too.
Did you consider using a WiFi module? I have tested a module with ESP8266. It worked a treat. Rumours are it reaches more than 100 m; a guy used the same module to remotely switch a water pump outside in his orchard. It needs a WiFi AP but no internet connectivity. The other module can drive the garage gate. The antenna is etched on the PCB. I tested something like this:
https://www.banggood.com/ESP8266-ESP-12E-Remote-Serial-Port-WIFI-Transceiver-Wireless-Module-p-980984.html?rmmds=search&cur_warehouse=CN

VBR from Ivan

i will try that smovpo antenna at the rx next week, but the tx on the bentley  remains like it is because it's best positioning now. The rf meter tells me that is working. Every time i operate the tx it lights up even if the gate opens or not so only range issues can be found there.(i carry the rf meter in the car to see what is happening) 
Using internet switch is last option because i don't want to relay on gsm network.
I have a measuring device wich i used to position a parabolic dish. It is scaled in db and also beeps. Can i use that somehow to measure if i have the 6 db?

And there is something fishy in the article http://sm0vpo.altervista.org/antennas/6dbvhf0.htm
First statement : "Each element and the stub are all a calculated 1/2-wave in length"  
after a few rows below a second statement : "The only important bit is that the stub should be 1/4-wave in length, not counting the shorted (folded) end." 
Now is it 1/2 or 1/4? 
Watching the pictures the stub does look like 1/2

"My car is similar to that Bentley car just that i don't have an engine and four wheels."    
It is often not a good idea to omit essential parts of a working construction.
Please do not feel injured.

VBR from Ivan

not trying to put a parabolic dish 
My antenna is similar to that smovpo antenna just that i don't have matching stub and coax balun and i use it horizontally . 
I might reinvent the unknown 
Somehow i will get my 150 m one way or another .

zsolt wrote:what if i cut a small pipe long as my dipole antenna and cut it in half along it's axis. If i put the dipole in the half pipe shouldn't that reflect all signal in forward direction?How comes that parabolic dishes reflect signal instead of stealing it?
I will not use the car's antenna. To complicated.
I stick to the dipole.
Would it help if i make the receiver dipole antenna full length?
For the car an other possible location would be on the rooftop trunk cross bars. This could be also full wave but it would be not more then 15 cm away from the roof.

Please, read some literature about UHF antennas an do not try to invent what is already known.
Antennas with reflector have been invented. The reflector is a bit longer than half wavelength, often in one piece. It may be grounded in the middle, as there is a node of voltage there. The distance between the radiator and a reflector is rather critical. (Added some shorter elements - directors - on the opposite side, a Yagi antenna is what you get.)
A parabolic dish for 433 MHz is a real overkill in your case. It would be several meters in diameter.
Two collinear elements half wavelength each? It is possible, but the phasing and matching elements may need thorough tweaking. This may help you: http://sm0vpo.altervista.org/antennas/6dbvhf0.htm   Scale it for 433 MHz.
A horizontal antenna 15 cm above the roof? Give it a test. I am afraid its performance will not be much better.

VBR from Ivan

what if i cut a small pipe long as my dipole antenna and cut it in half along it's axis. If i put the dipole in the half pipe shouldn't that reflect all signal in forward direction?How comes that parabolic dishes reflect signal instead of stealing it?
I will not use the car's antenna. To complicated.
I stick to the dipole.
Would it help if i make the receiver dipole antenna full length?
For the car an other possible location would be on the rooftop trunk cross bars. This could be also full wave but it would be not more then 15 cm away from the roof.

It's more easy to put outside tx antenna. To use the existing antenna on car roof i think i should chop of the tip of it so only 34 cm will remain. Also i think a small rellay is needed to switch the antenna from the radio to the tx.Probably the radio will steal power. Then i need to change the antenna from the garage roof exactly like the one from the car. So both antennas would be vertical and on top of roof's. What about that?
The other possibility is to move the dipole i built behind the rear view mirror somewhere outside. If i put it on the front bar in front of the radiator it's same thing i believe, the car would still absorb rf power. Plus it would look like the car has a mustache.
Ps.today the remote worked from 30m. It's sunny and no wind.
Now i don't know what to believe but i have suspicions about the rx antenna. Sometimes working and sometimes not seems like a contact problem but i did not find any

Hi,
12 m is quite enough if the TX antenna is inside a car. If both antennas are in free space, far (more than 10 lambda i.e. 7 m) from conductive bodies and there are no obstacles between them, the range may be up to 100 m or even more. Fog naturally increases the atmospheric absorption, decreasing the range.
You will not be happy with possible solutions: use a more sensitive receiver (limited by external, not own noise), transmitter power up to the legal limit for license free ISM, directive antennas with as much gain as you can afford and maybe math processing of the received signal to receive the information below the noise level.

VBR from Ivan

hi 
the rf meter seems to be useful so far. 
I modified code so that the Tx is always on . I have 60 mv @ 50 cm away and about 20 mV @ 1 m away from tx antenna. Beyond that the meter is not able to show something . 
Now what i do to increase range?
I observed that when is fogg outside the system is not working at all. When is clear i can open the gate ~12 m away

Hi Zsolt,
I would try a stub of coax (or maybe a whole comb structure of stubs) as a filter. The indicator itself is really wideband, ranging from LW to microwaves.

VBR from Ivan

P.S.: Do not worry that the device looks ugly. Get it to work first, tidy it up later (or maybe never )!

I did a little modification to the display after studying the LM3914 datasheet and a lot of absorption rf meters from the net. 
A lot of the schematics use 100 ....250 uA needle instruments . In analogy with that i modified the LM3914 circuit that i found to measure from 0 to 114 mV.
For this i modified  the pcb ... now the whole circuit looks so ugly that it hurts eyes , but it seems to work 
The schematicus i ended up is pictured like this:

I say it seems to work because meantime during a phone call i saw (almost scared ) that all LED's light up and flicker.
I found that i can even trace wires in the wall with this thing. Tomorrow i will make some time to see  what is happening with the TX from the car. So far so good.

Since  this circuit seems to react to everything i would make it to react to 433 MHz only . Some schematics where using some kind of resonator instead of the diode which goes to ground. I do have 433Mhz cristal from an other transmitter
The eye hurting product looks like this

looking at the LM3914 block diagram i see a 20k rezistor on the signal input. From the whole document i can't find what is the input impedance. With this i will deal later. I plan to use 2 glass case diods. If not working i will try schotky. Meanwhile still searching for parts

I did a little test with 2 cooper rods as antenna.
Of course it des not work.
If i take away the potentiometer completely the leds light up when i get close to a monitor or even the soldering iron

The LM3914 is more like a voltmeter measuring the voltage drop on the potmeter . On the original rf meter the indicator is directly an ampermeter. Maybe that's the difference.
If i put a DVM on the potentiometer i get 40 mV when i approach the thing to the monitor. Maybe an amplifier is needed . Or the needle instrument needs to go..
Anyway now i know that the transmitter is rally week.

zsolt wrote:Hi,
i started building the rf meter ....

Probably no microammeter has 500 Kohm internal resistance. I expect somewhere about 1 Kohm. The trimmer in your device should have similar value.

VBR from Ivan

They are pretty old. The upper right is almost surely a germanium one. Germanium diodes have forward voltage drop about 0,3 V only. Check them with your meter.

VBR from Ivan

i think these are the oldest diode s i will ever find

Yes, that looks nice.

VBR from Ivan