NS73M FM radio transmitter code

[NXTreme]

I recently purchased a NS73M FM radio transmitter module from SparkFun. I'm planning to put it into a project enclosure along with a Picaxe 20X2, a serial LCD and a couple buttons, all to create a small, portable FM radio transmitter. However, I still not sure that a Picaxe would actually be able to control the transmitter. I found an Arduino program for this module but it raised a couple questions about what the Picaxe can handle. First off, the program (http://www.km5z.us/files/arduino-fm/ARRduino_FM_v1/ARRduino_FM.pde with the project detailed at http://www.mikeyancey.com/FM-Stereo-Broadcaster.php) uses numbers that are bigger than a word can handle. I think this could be solved with a large enough lookup table, which the 20X2 could probably handle. Also, my understanding, or abilities, to read C are not very good, so I may be missing some other commands that could present problems.

Has anyone else here ever experimented with this FM transmitter module? If you have I'd love to know if it's at least possible to get it working. Or, just a hint or two to push me down the right path. One quick note. As most would already know, transmitters like these are illegal in some countries if you use much more than a short piece of wire as an antenna. Fortunately, I live in a desert, surrounded by mountains on all sides, and not in the US. So, I don't have to worry too much about disrupting either of the two radio stations that I can receive!

 

[srnet]

A clue as to its intended use would help ..... portable FM transmitter for what ?

Its an I2C device so the 20X2 should drive it. You should be able to work out which registers to fill with which values by calculations outside of the PICAXE and then just write them out to the chip, the PICAXE does not necessarily need to do any calculations at all.

But then it rather depends on what you want to do with the device ......

 

[NXTreme]

Ahh yes, that could be helpful. My initial use of it will probabhly be as a drive-in movie theater audio transmitter. It would be set up as a fundraiser for the students of a small, private elementary and high school. They raise funds throughout the year and at the end of the school year, go on an extended field trip. However, after being used for a day or two at the "movie theater" I would probably use it to transmit audio throughout the house. This means that it would most likely contain one audio input, a 3.5mm stereo jack for MP3 players as well as some additional circuitry to step down line in voltages to the 200mv RMS that is the upper limit on what can safely be applied to the NS73M module.

I'm not sure if this is what you needed to know but I'm willing to provide more information if need be. Thanks for your interest!

 

[srnet]

If its only to be used on one frequency then do the calculations in something like Excel, and just load the required values into the chip with the PICAXE.

In fact just like the 'Example Code in Basic' link on the Sparkfun link you gave us, although its not in PICAXE basic it gives you the idea, did you look at that code ?

 

[hippy]

Looking at the code, the frequencies used are ...

#define topFM 107900000 // Top of the FM Dial Range in USA
#define botFM 87500000 // Bottom of the FM Dial Range in USA

Too big to hold in a PICAXE 16-bit word as you say, but when we look at the SaveFrequency() and LoadFrequency() routines only a 16-bit word is stored with a divide and multiply by 10000, so the actual valid frequency numerical values are 8750 through 10790 which the PICAXE can handle.

The full numbers are simply convenience in handling the frequency in Hz but no reason not to handle them in multiples of 10kHz and simply add zeroes for display ...

For freq = 8750 To 10790
SerTxd( "Frequency = ", #freq, "0000 Hz", CR, LF )
Next

It's easy enough to adjust that so 8751 displays as "87510000 Hz", "87510.000 kHz" or ""87.510000 MHz" and drop trailing zeroes as desired.

I'm not familiar with the NS73M but from the datasheet and code it seems there is a 16-bit frequency value to set and an 8-bit band register ...

Band 3 when < 8850
Band 2 when < 9790
Band 1 when < 10300
Band 0 otherwise

For the 16-bit frequency value written to the chip, that is ...

( MHz + 304000 ) / 8192

So for our 'freq' value in steps of 10000Hz ....

( ( freq * 10000 ) + 304000 ) / 8192

or

( ( freq * 80000 ) + 2432000 ) / 65536

That should be possible with PICAXE maths but someone other than me is going to have to work out the fine details

So perfectly doable with a PICAXE, and 'big numbers' not really much of a problem.

 

[Armp]

I'll take a shot at it Hippy.

FM frequencies are usually specified with only 1 decimal place. So the frequency range is 87.5 to 107.9 Mhz and can be represented as Freq = (875 to 1079)/10

Equation from the data sheet is

Div = (FreqMhz+.304)/.008192​

Bit of evaluating constants, and rearranging to avoid overflow gives:

Div = (Freq - 875) *122/10 + 10718 where 875<=Freq<=1079​

Example from datasheet is 88.5 Mhz -> 10840. We're close enough.

As for the intended application - this Tx only gives out 2mW - you're not going to get much range.

 

[srnet]

First off, the program (http://www.km5z.us/files/arduino-fm/ARRduino_FM_v1/ARRduino_FM.pde with the project detailed at http://www.mikeyancey.com/FM-Stereo-Broadcaster.php) uses numbers that are bigger than a word can handle

What are you trying to do exactly ?

Do you want to program it as a fixed frequency transmitter ?

or

Do you want to be able to adjust the frequency from the PICAXE ?

If the latter, then what you likely need is a routine (in PICAXE basic) to display the frequency set according to the current values of the registers.

 

[Grogster]

As for the intended application - this Tx only gives out 2mW - you're not going to get much range.

I have to agree with that.

For all intents and purposes, 2mW is useless as an audio transmitter - you could expect about 50 meters range TOPS, LOS(Line of sight, NOT through walls - if walls in the way, probably about 20 meters)

A low power option might be this one: US$28.50
http://www.mdfly.com/index.php?main_page=product_info&cPath=8_50&products_id=143

Or this one with LED display on-board:
http://www.mdfly.com/index.php?main_page=product_info&cPath=8_50&products_id=142

Or this fully boxed 500mW unit, which I have used several of and are exceptionally good for the cash.
http://www.mdfly.com/index.php?main_page=product_info&cPath=8_50&products_id=454

The half-watt job will exceed FCC regulations, and many other places in the world too probably, but here in New Zealand, we are allowed 1W on LPFM licence-free, so I actually use on of these with a Free Radio Berkeley blower amp on it to give me the extra juice.

 

[NXTreme]

Sorry for the delayed answer. I think that with the great information you all have provided I shall be able to get this working. Thanks for all the replies!

If its only to be used on one frequency then do the calculations in something like Excel, and just load the required values into the chip with the PICAXE.

In fact just like the 'Example Code in Basic' link on the Sparkfun link you gave us, although its not in PICAXE basic it gives you the idea, did you look at that code ?

I want to be able to change the frequency on which it transmits, hence the LCD and buttons. I think that your idea of calculating the values beforehand and entering them into the Picaxe is a good one. I could probably use some sort of lookup table to accomplish this.

That should be possible with PICAXE maths but someone other than me is going to have to work out the fine details

So perfectly doable with a PICAXE, and 'big numbers' not really much of a problem.

Good to hear that. I think that with the math you gave me I can figure out the rest on my own, barring any unforeseen problems.

As for the intended application - this Tx only gives out 2mW - you're not going to get much range.

Thank you for expanding on the math stuff, appreciate it. Yes, I know that 2mW isn't going to go very far. However, I figure that with a nice, big antenna I could get at least 50 meters range, which is enough for my purposes.

What are you trying to do exactly ?

If the latter, then what you likely need is a routine (in PICAXE basic) to display the frequency set according to the current values of the registers.

Yes, the latter is what I'm going for. Now that I have more of an idea on how to program the transmitter, the rest is easy. I have the LCD and button control code written up already.

I have to agree with that.

For all intents and purposes, 2mW is useless as an audio transmitter - you could expect about 50 meters range TOPS, LOS(Line of sight, NOT through walls - if walls in the way, probably about 20 meters)

I agree with you, 2mW is not much for most purposes. However, I really only want a maximum of 50 meters range for now.I shall bookmark those links for future reference, just in case I want to start my own pirate radio station . Thank you for pointing out those other transmitters for me!

 

[nbw]

@Grogster: those 500mW units - you'd recommend them? In 2012 I'm toying with the idea of a low power FM radio station, maybe a few kms range. We're fairly high up on a hill so 0.5 - 1W might make it across the Wellington harbour

 

[Grogster]

@NXTreme - Oh, if you only want such a short range then... I thought a drive-in could be kinda big and was not sure 2mW would do it for you. Get your transmitting aerial exactly right - any mis-match at 2mW, and you won't get any radiated power at all

@nbw - PM me if you want to discuss this in more detail. There are lots of options out there for FM transmitters, but yes, those little boxed units are actually very good. I checked one on the spectrum analyzer for harmonics and other nasties, and they are actually very clean with harmonics and spurious emissions well suppressed. When I cracked one open to look at it, there is a complicated SMD RF filter arrangement on the output, so it's not just one of those rough-as-guts outputs. Considering the price of them, I doubt you would find any other stereo transmitter with similar 500mW output for the price.

I am looking at another 500mW boxed up unit for US$90. I have one coming as a sample, so I can let you know(PM you) once I have it if you like. Fully boxed with buttons on the front, and LCD display.

...but I digress from the topic of this thread...

 

[NXTreme]

@NXTreme - Oh, if you only want such a short range then... I thought a drive-in could be kinda big and was not sure 2mW would do it for you. Get your transmitting aerial exactly right - any mis-match at 2mW, and you won't get any radiated power at all

Well, normally a drive-in theater would be bigger, but it is a very small school and there probably won't be more than ten cars . About the antenna. What kind of antenna did you use for your transmitters? I would be making it myself, of course, but what kind of antenna do you suggest?

 

[NXTreme]

Sucess!

Well, it works! I finally spent a couple hours last night finishing everything up and making sure the wiring hadn't been messed up. I loaded the code onto the Picaxe 20X2, plugged in the FM transmitter module, power-cycled it and voila, it worked! I was actually surprised it worked on the first try, most of my projects never do . I've posted the code I used and a picture of the breadboard below. Keep in mind that this is just a simple test, no screen or other fancy peripherals attached, just the Picaxe and the FM transmitter transmitting at a fixed 93.9 MHz. I used an MP3 player as an audio source. The one small problem I encountered was a lack of volume. The audio was crystal clear out to about 70 meters, where it started to get fuzzy, but I had to turn up the volume almost all the way on my Android mini-tablet to hear it "loud and clear". The one thing I was thinking was that maybe my MP3 player isn't playing as loud as the FM transmitter needs the audio for a loud transmission. My 1960s Heathkit oscilloscope just died two days ago so now I have nothing with which to find out exactly how loud the MP3 player is playing. However, considering how much I've paid for this project so far (~$20 US) I'm pretty pleased with how it turned out.

As a warning, the code isn't mine! I translated it into Picaxe BASIC but the actual code itself was copied from here (http://mikeyancey.com/files/arduino-fm/ARRduino_FM_v3/ARRduino_FM_v3.pde).

	#picaxe20X2

init:	hi2csetup i2cmaster,%11001100,i2cslow_8,i2cbyte

	pause 50

main:	hi2cout 14,(%00000101) 'Software reset
	pause 1
	hi2cout 1,(%10110100) 'Pilot on, forced subcarrier
	pause 1
	hi2cout 2,(%00000011) '2 mW power, unlock detect off
	pause 1
	hi2cout 3,(%11101010) 'Lower byte freq. 93.9 MHz
	pause 1
	hi2cout 4,(%00101100) 'Upper byte freq. 93.9 MHz
	pause 1
	hi2cout 8,(%00011010) 'CEX=Band 2
	pause 1
	hi2cout 0,(%10100001) 'Pwr on, 200 mV RMS, 75 uS pre-emph
	pause 1
	hi2cout 14,(%00000101) 'Software reset
	pause 1
	hi2cout 6,(%00011110) 'Set internal charge pumps
	
	end

 

[Grogster]

Well, normally a drive-in theater would be bigger, but it is a very small school and there probably won't be more than ten cars . About the antenna. What kind of antenna did you use for your transmitters? I would be making it myself, of course, but what kind of antenna do you suggest?

Basically, I would go for quarter-wave or half-wave length of stiff wire. It will work OK no matter what wire you put on it, but with only 2mW to play with, a correct frequency-matched length will get you the best possible range.

The length of the wire depends on weather you elect to use 1/4 wave or 1/2 wave, and the frequency you are transmitting on.

This link will work it out for you, then all you have to do, is cut the wire to the correct length:
http://www.csgnetwork.com/freqwavelengthcalc.html

Scroll down a bit to get to the calculator.

ADDITIONAL: As a reply to your post #13 above: WELL DONE!!! I will be watching this thread for your future development of the concept!!!

Quarter wavelength @ 93.9MHz = 798.7mm or 1597.4mm at half wavelength.
For best range, use half-wavelength, but due to room constraints, you may have to stick to quarter wavelength. Aerial wire must be vertical - do not bend it to make it fit inside the room, or run it around the edge of the ceiling. DO NOT place the aerial within about a meter of anything metal, as metal close to the aerial will affect range. Best possible location is sticking up out the top of the roof, but for your purposes, taping the wire vertically to the wall(away from any metal) should suffice.

On the lack of volume, most MP3 players and tablets/phones don't have much output(which is probably a good thing for teenager's ears!), so you could try the line-out from a DVD player, which will be a much higher output level. The module datasheet should specify what the audio input level should be.

 

[NXTreme]

Basically, I would go for quarter-wave or half-wave length of stiff wire. It will work OK no matter what wire you put on it, but with only 2mW to play with, a correct frequency-matched length will get you the best possible range.

Ok, I'll try that. I was using a piece of wire that was cut to approximately 1/4 wave at 93.9 MHz but it wasn't straight, nor very exactly cut. I found when experimenting with my 315 MHz data transmitter modules that a 1/2 wave antenna worked much better than a 1/4 wave, however that might not apply in this case. I have a couple meters of thick copper wire that I will use, will most likely hold it's shape better than the other wire I have.

ADDITIONAL: As a reply to your post #13 above: WELL DONE!!! I will be watching this thread for your future development of the concept!!!

Thank you! I have a small metal Palm Pilot case that I will try to fit everything in. The one problem I ran into was that the only serial LCD I have runs at 5V, not 3.3V like everything else. To properly run everything I will use a two cell LiPo battery along with a 5V and 3.3V LDO regulator. I also bought some TV coaxial cable and connectors that I might use if I need to position the antenna higher up than I want to put the transmitter, Picaxe and screen.

For best range, use half-wavelength, but due to room constraints, you may have to stick to quarter wavelength. Aerial wire must be vertical - do not bend it to make it fit inside the room, or run it around the edge of the ceiling. DO NOT place the aerial within about a meter of anything metal, as metal close to the aerial will affect range. Best possible location is sticking up out the top of the roof, but for your purposes, taping the wire vertically to the wall(away from any metal) should suffice.

I have plenty of room inside the house, so the size of the antenna shouldn't be a big concern. The one problem is that the house is made of bricks, concrete and rebar. This means that I would probably have to hang the antenna from the ceiling or mount it on a pole extending from the wall. The range doesn't seem to suffer too much however, as I've been able to receive clear audio all over the house without hardly any hiss.

On the lack of volume, most MP3 players and tablets/phones don't have much output(which is probably a good thing for teenager's ears!), so you could try the line-out from a DVD player, which will be a much higher output level. The module datasheet should specify what the audio input level should be.

The NS73M datasheet states that 200mV is the max audio level input. From what I remember 200mV is about what an MP3 player/phone will put out at it's loudest. However, I don't want to burn out the transmitter . I might be able to use the local computer techs oscilloscope in a couple days, will have to ask him.

Thanks for all the help! I would not have been able to get this far without you guys!

 

[Grogster]

I have plenty of room inside the house, so the size of the antenna shouldn't be a big concern. The one problem is that the house is made of bricks, concrete and rebar. This means that I would probably have to hang the antenna from the ceiling or mount it on a pole extending from the wall. The range doesn't seem to suffer too much however, as I've been able to receive clear audio all over the house without hardly any hiss.

Yeah, so is my house, and it has a corrugated iron roof which acts like a giant RF shield to anything I test-transmit from inside!
Nothing is perfect, but as much as practical, keep it away from metal - I like your idea of hanging it from the ceiling on a hook - that would keep it away from any other metal thing. You could even hang it from the ceiling light, so long as the light shade was glass or plastic...

The only reason I am being so pedantic with respect to the aerial, is that you don't have a lot of juice to play with. If you had a few hundred Milli-watts, you could get away with a less efficient aerial, as although you may lose some power in the not-perfect match, there is enough reserve juice to still transmit just fine. With 2mW output and a wrong aerial wire length, you basically will lose all the output power to the mis-match!

But generally speaking, if you think I am being to technical, then just about any bit of wire will work, and you can ignore me!

The NS73M datasheet states that 200mV is the max audio level input. From what I remember 200mV is about what an MP3 player/phone will put out at it's loudest. However, I don't want to burn out the transmitter . I might be able to use the local computer techs oscilloscope in a couple days, will have to ask him.

Thanks for all the help! I would not have been able to get this far without you guys!

Yeah, in that case, the line-out from a DVD player probably would overload the module's audio input, so probably not such a great idea of mine to do that after-all!!!

 

[srnet]

If the radio module is designed for a circa 50ohm impedance antenna, as in a end fed 1/4 wave, then a similar half wave antenna will be a terible match and should make things worse.

However I suspect its a waste of time cutting an antenna to an exact length, best to find the optimum by experiment.

 

[Grogster]

Good point - I don't have the datasheet, so I will leave that with NXTreme...

However I suspect its a waste of time cutting an antenna to an exact length, best to find the optimum by experiment.

I was actually starting to think the same thing myself, in that it might not be worth the effort:

But generally speaking, if you think I am being to technical, then just about any bit of wire will work, and you can ignore me!

 

[NXTreme]

But generally speaking, if you think I am being to technical, then just about any bit of wire will work, and you can ignore me!

Not at all! I love technical! That's why I took up robotics/electronics and why I love tinkering with almost anything that is electronic/mechanical. I may not always understand something but that usually gives me a reason to start searching for more information about what I don't understand. You could say that by being technical you are making me smarter . I want to have the best antenna possible (within reason, as I can only create certain things with the limited materials available to me) for the exact reason you mentioned, that this transmitter is severely under powered. I have been doing some Googling of my own and am thinking that if I have time over Easter, I might create a J-pole antenna, just to see if I can get any extra range out of it.

If the radio module is designed for a circa 50ohm impedance antenna, as in a end fed 1/4 wave, then a similar half wave antenna will be a terible match and should make things worse.

However I suspect its a waste of time cutting an antenna to an exact length, best to find the optimum by experiment.

The datasheet (http://www.sparkfun.com/datasheets/Wireless/General/NS73_Datasheet.pdf, page five) states that it is designed for a 50 ohm antenna. I don't know terribly much about RF design and all that so I don't know of what significance that is, but I'm open to suggestions. The one thing I was thinking of was that I could, instead of experimenting with antenna length, experiment with transmitted frequency. Almost the entire FM radio band is open for me to use so I can easily experiment without interference. This would (as far as I know) have the same effect as changing the antenna length and be more quickly done.

 

[srnet]

The datasheet (http://www.sparkfun.com/datasheets/Wireless/General/NS73_Datasheet.pdf, page five) states that it is designed for a 50 ohm antenna. I don't know terribly much about RF design and all that so I don't know of what significance that is, but I'm open to suggestions. The one thing I was thinking of was that I could, instead of experimenting with antenna length, experiment with transmitted frequency. Almost the entire FM radio band is open for me to use so I can easily experiment without interference. This would (as far as I know) have the same effect as changing the antenna length and be more quickly done.

If its designed for 50ohm, then stick to a 1/4 wave vertical antenna, a 1/2 wave has completly the wrong impedance. At such low powers the mismatch will likley not cause damage to the module, but radiated power may be substantially less.

Tuning the antenna is not difficult. Start with a long length of plain wire, say somewhat over a 1/2 wave length. Measure the acceptable reception range and cut bits of the wire off and see the effect, better or worse. Its possible that at some length reception range will peak and then fall off as you remove more wire.

 

[manuka]

@Grogster: those 500mW units - you'd recommend them? In 2012 I'm toying with the idea of a low power FM radio station, maybe a few kms range. We're fairly high up on a hill so 0.5 - 1W might make it across the Wellington harbour

nbw: I live across this very harbour & can verify ½W from an elevated site should do OK. However decent reception GREATLY depends on the listeners FM receiver- some are VERY DEAF...

But just who is your target audience? There are now so many FM stations in most cities that it may be hard to justify studio overheads against the listener support. As an example a LPFM station I was once involved with decided on a whim to give a dozen beer to the first caller. They had absolutely no response - tellingly indicating that probably no one was actually listening...

You may get more support hence by directly broadcasting onto the web to a global audience.

 

[Grogster]

Yes, manuka has a point there.

A few years ago, FM radio stations were cool, and pirate ones were even cooler!

However, two things have happened since then: (1) A very crowded FM band - it can be hard to find a blank space to transmit in these days as the low-band FM and high-band FM allocations for LPFM have pretty much all been gobbled up(it's first come, first served for LPFM), and (2) As manuka hinted at, net-radio has much more potential simply because anyone anywhere in the world can tap into your stream and listen, which is way more range then you can get with a physical transmitter. Also, as there is no censorship on the web, you can literally say whatever you like, including as much profanity as you like, and you won't get into any trouble(unless you are really taking it too far), whereas with a physical transmitter, you have to behave yourself and adhere to the Broadcasting Standards Authority - bad language could easily get you shut down or fined or both.

I guess it is a personal choice, but DON'T be tempted to transmit outside of the LPFM band(88.1-88.7 and 106.7-107.7 in New Zealand) - I did that once(100.0), and was tracked and fined by RSM(Radio Spectrum Management). But I was transmitting 15W at the time, which was a little naughty...

 

[nbw]

I was quite keen on a smallish radio station, giving it a go (for a laugh?!) just to see if anyone did listen. Plus, I wanted to have or build some hardware. Computer is 2nd choice My very basic understanding of antennae etc is while there is a 1W broadcasting restriction transmitters, there isn't on aerials - so a decent bird perch potentially might get me quite far (maybe a couple of kms - I'm perched on a hill in Khandallah and can see the stadium (about a km away), the South Coast, Oriental Bay). I'm pretty good with my language until Jesse Ryder throws away his wicket. I'd stick to the LPFM allocations, maybe a bit of research would show which were the least congested.

@grogster: 15W - CRUMBS! Bet the RSM weren't too happy!
@manuka: I wouldn't mind giving web radio a go - any advice or suggested starter sites?

 

[Grogster]

I was quite keen on a smallish radio station, giving it a go (for a laugh?!) just to see if anyone did listen. Plus, I wanted to have or build some hardware. Computer is 2nd choice My very basic understanding of antennae etc is while there is a 1W broadcasting restriction transmitters, there isn't on aerials - so a decent bird perch potentially might get me quite far (maybe a couple of kms - I'm perched on a hill in Khandallah and can see the stadium (about a km away), the South Coast, Oriental Bay).

Yeah, A directional YAGI would push all your power out over the bay for maximum range in that direction, and you are correct - there is no restriction on the antenna gain(as far as I can remember). I think YAGI's a good for at least 2dB gain, but others here will be able to give you specifics.

@grogster: 15W - CRUMBS! Bet the RSM weren't too happy!

The radio inspector was just fine about it, but you are right - the RSM was not too happy, and that cost me $250, so YOU HAVE BEEN WARNED!!! My transmitter was compliant with no harmonics or other nasties, it's just I was exceeding the output power. That, and I should not have been running on that frequency either. I ran 15W for about 8 months though, before they found me...

 

[Grogster]

Bleepin' double-post again...

 

[nbw]

Hell eh. Seeing as I'm an old blighter (40+), I think I'll learn the rules before I break, er, adhere to them. So one of the 0.5W transmitters you've mentioned might be just the ticket, with a wee amplifier to take it up to a roaring 1 million micro-watts. And, a crash course in Yagi, provided the exercises are easy. Oh wait, that's Yoga.

 

[Mickeythetech]

Hi, As this is my first time posting to the forum please for give me if it's not right. I have been working with A NS73M break out board from Sparksfun. I'm using the code on post #13. With a Picaxe 08M2 on a picaxe 08 proto board. It's works when I first load the program. But when I power it off and on it will not load with out reloading the program again. I have tried several commands at the beginning and at the end of the program with no success. I read a lot about the Picaxe workings. But I can't find the answer to this question. Will i2c function work when the controller flash's the program. I have beeped out the proto board and it seems to be ok. Any suggestion would be deeply appreciated.

Thanks,

Mickey

 

[westaust55]

Hi, As this is my first time posting to the forum please for give me if it's not right. I have been working with A NS73M break out board from Sparksfun.

Mickey has also started a new thread in the Active Forum area:http://www.picaxeforum.co.uk/showthread.php?24240-I2C-function-at-flash-picaxe-08m2
To try and keep it all together, lets try to stay with providing answers in the other thread which was sighted and answered first.

 

[Paix]

Compared with a 1/4 wave whip 0dBi (isotropic), a simple dipole has a gain of around 2.15dBi. This is also known as 0dBd (dipole as the reference).
To be a Yagi it needs to have a driven element, a reflector and a director as minimum. I would be surprised if a figure of around 5dBd (7dBi) wouldn't be the minimum you might expect.

The useful range of gain is probably 5dBd to 12dBd. 6dBd to 7dBd will probably give you the shape that you want, without trying to take anyone's eye out and be relatively easily achievable

 

[MikeAusP]

Compared with a 1/4 wave whip 0dBi (isotropic), a simple dipole has a gain of around 2.15dBi. . . . .

A 1/4 wave whip is not an isotropic radiator - it's far from isotropic.

An Isotropic Radiator is a theoretical device that radiates equally in every direction - a bit like a star out in space.

A dipole has 2.15db more gain than an Isotropic Radiator - in it's plane of maximum radiation - because it focuses its energy into the plane perpendicular to its elements, resulting in a dumbbell shaped radiation pattern with zero radiation off the tip of its elements.

 

[Paix]

Thanks for the update Mike

I probably didn't state it very well, but essentially gain of dipole based antennas, such as Yagi-Uda, tend to be stated in dBd, whereas whips, colinears and the likes tend to be described in terms of dBi. The ARRL a few years ago took a dim view of traders specifiying the gain of Yagis citing the dBi gain, considering them to be engineering an unfair or unscrupulous psychological advantage.

The point that I was hoping to make was that there is a difference in the figures and that caution is advised to ensure comparing like with like.

@Mike you think of a star out in space and I think of an illuminated ping-pong ball. I wonder if anything that says about each of us . . . I guess that the star is more common example. I must check on the current ping-pong ball developments trends in China to see if my thinking stands a chance