HI guys i want to build an FM superhet receiver for my RC helicopter. Range will have to be about 200m. The transmitter is no problem thats already built. One problem though is the amount of infromation that has to be relayed to the reciever about 20,000 bits/second. I'm not asking for a circuit diagram or anything but i do have some questions:

Seeing that a superhet design revolves around the idea of mixing signals down to a much more manageable signal frequency, How exactly do the signals get mixed ?? Do i just connect the outputs from the RF tuner and the LO together and then feed them into an amplifier?? Someone told me that this was adding, not mixing. :?

Say the modulation bandwidth is -+5khz at 20mhz, after the mixing reduces this signal down to 455khz will the modulation bandwidth still be 5khz?? or this be reduced down as well. (ie, will it stay 450khz to 460khz)

Why do some recievers use 'dual-conversion' as in the RF signal heterodynes to form 10.7mhz, which is then heterodyned again to form a 455khz signal. cant you go straight from RF to 455khz, and just skip 10.7mhz??

I know there are license free modules available that could easily cover the requirements, but the prime reason in doing this is to learn about RF. Getting the model helicopter in the air is just a bonus.

edit: And why do AM receivers sometimes have 3 IF filters all based on 455khz, couldn't you use just one filter?? Does this have something to do about the Q of the IF tuners and the selectivity??

__________________
www.winpicprog.co.uk - Great PIC language tutorials.

 

They mix in a mixer (funnily enough!), but not in the same way as an audio mixer (which is actually an adder). There are numerous different types, but basically double-balanced, single-balanced, and non-balanced, with double-balanced being the best.

A mixer provides four outputs:

1) Input one

2) Input two

3) Difference between them (1-2 or 2-1)

4) Sum of them (1+2)

You want either 3) or 4), the selected one is decided by the IF frequency of the following stages.

A non-balanced mixer will output all four, and rely on the filtering to reject the rest (almost all domestic radios are like this).

A single-balanced mixer will balance out one of the inputs, providing only 3), 4) and either 1) or 2), depending which is balanced.

A double-balanced mixer balances out both inputs, so you only get 3) and 4) output.

Essentially a mixer is a non-linear device, something as simple as a single diode will do the job, but more commonly a single transistor or IC is used. A popular component is a dual-gate MOSFET, with each input applied to one of the gates.

It doesn't change, that's the whole point of it, so your +/-5KHz will still be that at 455KHz.

To give better selectivity, narrower bandwidth, and to better reject image frequencies.

Going back to the four frequencies earlier, assume we are tuned to 100MHz, with the IF at 10MHz? (easy maths value :lol: ), the local oscillator could be 90MHz, so 90-10 gives 80MHz (not wanted), and 90+10 gives 100MHz (wanted).

So TWO input frequencies can be picked up, 100MHz which we want, and 80MHz (the image) which we don't. The front end tuning of the radio needs to be sufficient to reject the image, while allowing the required one through.

If your IF is only 0.5MHz (easy value again :lol: ), then the two possible inputs become 90.5MHz and 89.5MHz (assuming the same 90MHz local oscillator) - the front end tuning is VERY unlikely to be able to be able to accept one and reject the other. This is why you have a high IF frequency, to provide an acceptable image frequency.

The reason for the second stage (455KHz) is to give good selectivity, a 10.7MHz IF is wide band, down-converting to 455KHz gives much narrower bandwidth and good selectivity - the best of both worlds!.

I don't know what the legal situation in Australia is?, but in the UK aircraft model radio systems are covered by very strict legislation, they have their own specific band, channels and modulations.

Bear in mind your helicopter is an EXTREMELY dangerous weapon, and could very easily kill someone. Assuming that happened?, you would quite rightly be guilty of murder!, and presumably would be sentenced accordingly.

I would STRONGLY suggest that you look in to the legal implications of this! - also the licence free radio modules would NOT be legal for this purpose!.

I would also strongly suggest you buy a commercial aircraft radio system, if you want to experiment with radio, DON'T do it with an aircraft.

To give better selectivity, with simple transformers the selectivity is a LOT less than a ceramic filter, so you use one in each IF stage, and you can stagger tune them to set the actual bandwidth (within certain limits). It even improves performance by adding extra ceramic filters as well.

As with all radio questions though, you should buy a copy of either the RSGB or ARRL handbooks - even for non-radio questions, EVERYONE on these forums should own at least one (if not both).

__________________
PIC programmer software, and PIC Tutorials at:
http://www.winpicprog.co.uk

 

Well according to www.aca.gov.au

So that's that sorted out. How far will that 300mw limit get me??
Nothing was said specifically about aircraft models and transmitter limitations, apart from the fact that the engine/motor cant have more than 4kw of power behind it. model cars and boats are allowed 8kw.

And according to that rule, my remote control Tank based on 433mhz licence free modules is illegal.

Do I have to use the standard 10.7 and 455khz frequencies or can i select my own IF to work with??

__________________
www.winpicprog.co.uk - Great PIC language tutorials.

 

There's no point having a range greater than you can see!, commercial radio control systems have a large enough range for any possible requirement.

No body ever suggested it wasn't illegal? - but there's a HUGE difference between a small model tank crawling about and an aircraft doing 80mph with sufficient energy to kill or maim a crowd of bystanders.

You could select your own, but why would you want to?, 10.7MHz and 455KHz are standard values, freely and cheaply available, there are a few other ceramic filter values available, but they are uncommon and expensive.

__________________
PIC programmer software, and PIC Tutorials at:
http://www.winpicprog.co.uk

 

I think the people who make the flying boat and flying lawnmower kits also make a flying tank. Or maybe it is a flying fire-truck?

My lawyer would love to have me hit by an uninsured flying tank! :cry: Aren't you rich? :lol:

__________________
Uncle $crooge

 

The mixing requires 2 inputs, A local oscillator input and an RF input. Do both signals have to be of equal power (as in voltage amplitude). Do they have to be the same. Does this mean I will require an automatic gain control circuit prior to mixing??

__________________
www.winpicprog.co.uk - Great PIC language tutorials.

 

No, they don't have to be the same level, and you don't apply AGC before the 1st mixer (apart from in extremely high signal strength conditions, where an RF attenuator can be switched in).

__________________
PIC programmer software, and PIC Tutorials at:
http://www.winpicprog.co.uk

 

Seeing that this circuit will be in CW mode to transmit digital information, Could I use a simple slope detector??

Some sources are saying that this type of detector is also prone to AM, the only way to counter this problem is to remove the AM all together using a limiter. Could this limiter be something as simple as say a zener diode???

Lastly, if I use that zener diode-limiter that means the tops of the sine waves are going to look "clipped", hence it will look more like a square wave. Will the slope detector work in such a condition, or do i have to keep it pure sine wave??? Is this where an AGC might come in??

EDIT: or could i just amplify it into clipping??

__________________
www.winpicprog.co.uk - Great PIC language tutorials.

 

A slope detector is a poor way for an AM detector to try to receive FM modulation.
A zener diode has a high capacitance and would attenuate RF.
An amplifier that clips has its transistors in saturation, and since it takes time for them to come out of saturation then their high frequency response is reduced and their output is delayed. Don't re-invent a limiting amplifier, buy a good one in an IC.

__________________
Uncle $crooge

 

So the detector will work with a square wave??

At 455khz, will that saturation problem matter much?? How about mosfets instead??

__________________
www.winpicprog.co.uk - Great PIC language tutorials.

 

Why don't you just use an FM IF IC?, there are a number of them (usually Motorola), designed for this exact purpose - dual conversion NBFM.

But if you're using FM, use a PROPER FM detector, don't try and bodge slope detection - I can see the death rate in Australia climbing even higher :lol:

__________________
PIC programmer software, and PIC Tutorials at:
http://www.winpicprog.co.uk

 

Note taken* Sorry i haven't posted in a while, i went straight to work looking for a chip. MC3361 Is the only available one. The chip handles with the limiting, mixing for the 455khz, demodulation, and second conversion oscillator (455khz).

The datasheet says i need a quad coil for the 455khz filtering. Can someone explain what a quadcoil is or does??

Heres the remaining circuit that i pulled together from various sources. All the chip needs is an IF input of 10.7mhz with the modulation: Criticise it!!

EDIT: Those funky colored boxes are rewound IF coils. The resonant frequencies are on top of them.

Attached Images

rx.jpg (25.8 KB, 686 views)

__________________
www.winpicprog.co.uk - Great PIC language tutorials.

 

Does the oscillator work?.

Why use an untuned preamp in front of the main preamp?, if you want to use two stages tune the first one as well - it's also a good idea to have a tuned circuit direct from the aerial, to help keep out of band signals out. But to be honest, I don't see the need for the extra amp in-front, just use the single tuned preamp, but with a tuned transformer on the input as well.

Why the 470 ohm in the emitter of the main preamp?, as it's un-bypassed it will provide large amounts of negative feedback - bypass it with a capacitor.

BTW, a 'quadcoil' is probably a 'quadrature coil', part of the FM demodulator.

__________________
PIC programmer software, and PIC Tutorials at:
http://www.winpicprog.co.uk

 

Pike,

The oscillator looks a bit iffy, also you would be better using a crystal oscillator it will give better stability than a free running LC oscillator.

Get rid of the first RF amplifier, it wont do you any favours, and use a tuned circuit with a link winding to couple in the antenna.

Why do you decscribe the 10.7Mhz IF transformer as "modified"? I would have thought that could have just been a standard 10.7Mhz item.

The collector of the RF amplifier should go to the tap on the primary of the 36Mhz transformer, otherwise the transistor will load the coil and the Q will be so low it wont have a resonance peak.

Same with the 10.7Mhz transformer, and you dont want the 10k resistor across it. It will just load it and flatten the response.

JimB

__________________
Experience is directly proportional to the value of the equipment ruined.

 

Notes taken*

That IF coil is modified to 10.7mhz, because I happen to have about 40 identical 455khz that i dont want to go to waste. That missing bypass capacitor was an simple editing error.

The oscillator should work. I pulled it out of the circuit audioguru designed here, the mod4 version that is: http://www.electro-tech-online.com/v...ic.php?t=14899. If it works at vhf I cant see why it wont work at 25.3mhz. The first diagram I drew looks iffy, probably because the resistor on the base was supposed to be a capacitor.

To save space, I decided to go with a simple prewound choke and capacitor in a series resonant filter setup. The reactance of the capacitor at 36mhz is about 550 ohms where as for the inductor it is about 490. Will this 60 ohms difference matter??

How does the circuit look now??
EDIT: that circuit diagram looks rather, well...naked. :?

Attached Images

rx-mk2.jpg (23.4 KB, 645 views)

__________________
www.winpicprog.co.uk - Great PIC language tutorials.