Question about iPod type exciters
Jack
Or are there possibilities here?
Jack
Bob
> Total crap for use as first stage of a transmitter?
>
> Or are there possibilities here?
>
Generally not worth bothering with - they use the simple "stereo
transmitter" ICs - these things generate huge numbers of spurs and
harmonics, and aren't really suitable to be "cleaned up".
These ICs typically produce a milliwatt or two, so would require significant
amplification to be in any way useful. They produce lots of "close-in"
spurs, so amplifying them will also amplify all the unwanted rubbish!
There are lots of simple but effective stereo coder designs on the 'net, and
an exciter isn't difficult to build. A stereo coder (mostly) works at
audio frequencies, so ordinary strip board, matrix board, PCB or "ugly"
construction will work well. However, if you have no experience of VHF
radio construction, you should consider a good kit for the exciter.
Worthwhile products are available from (in no particular order):
www.nrgkits.com
www.broadcastwarehouse.com
www.pcs-electronics.com
There /are/ other kits on the market, but they generally aren't worth
wasting your money on. You /really/ do get what you pay for!
Each of the above will provide you with a clean and stable output from 1 to
about 5 Watts according to product. You may find that this is enough.
However, if you find you need more power, any of the above provide power
amplifier kits or modules.
You'll also need to "process" your audio (with a limiter at least) to
prevent interference to adjacent stations.
Do some serious research, and learn the technical principles involved. Be a
responsible broadcaster, and don't cause interference or transmit offensive
material!
*Good* *luck!*
Bob
WSQT
a good enough engineer to build your own exciter!
A "premix" exciter with a low frequency(5-8 mhz) modulated VFO and a
high frequency crystal oscillator(Old CB crstal on 9th overtone, third
plus tripler if won't go on 9th) will be cheaper to build and probably
easier to filter, as you can design it so all the spurs are several MHZ
away. They can be filtered out to better than 70 db down by using 3-4
cascaded stages of doubly tuned 2N3904 low level amplifier stages.
Jack
engineering skills, but I operated a ham radio as a kid and a MARS station later in the army. I
built a couple of CW transmitters and designed and built a simple beam antenna so I think I
might be able to find my way through the task of building from schematics. I have found some
good sites and I'm sure I can find more, but if you know of any, I would certainly appreciate a
push in their direction.
I am new to pirate fm, but from what I have been able to determine two important factors are
stability and a clean rf signal. I would also like stereo and good audio quality. I don't need a
lot of versatility in terms of ease of changing the transmitting frequency. Twelve volt, plus or
minus, power input would be nice, but not absolutely necessary. First stage amplification would
be fine. I am mainly interested in the exciter stage. An amp would be easier to come by later,
but anything between .1 and 10 or 20 watts would be worth a look at.
All the best,
Jack
Bob
> I am looking into building something if I can find some good schematics. I
> don't have any engineering skills, but I operated a ham radio as a kid and
> a MARS station later in the army. I built a couple of CW transmitters and
> designed and built a simple beam antenna so I think I might be able to
> find my way through the task of building from schematics. I have found
> some good sites and I'm sure I can find more, but if you know of any, I
> would certainly appreciate a push in their direction.
>
> I am new to pirate fm, but from what I have been able to determine two
> important factors are stability and a clean rf signal. I would also like
> stereo and good audio quality. I don't need a lot of versatility in terms
> of ease of changing the transmitting frequency. Twelve volt, plus or
> minus, power input would be nice, but not absolutely necessary. First
> stage amplification would be fine. I am mainly interested in the exciter
> stage. An amp would be easier to come by later, but anything between .1
> and 10 or 20 watts would be worth a look at.
It's /really/ a waste of time trying to filter out the mess from one of
those "yard-caster" ICs. It /might/ be possible, with several stages of
critical filtering as WSQT suggests, but you're /much/ better off starting
with a clean source to begin with!
It sounds like you have the basic electronic skills necessary to build your
own. WSQT suggests the use of a "pre-mix" approach - again this requires
fairly critical filtering to prevent unwanted spurs, and /really/ needs the
use of a Spectrum Analyser to set up accurately - you probably don't have
one to hand!
If you're going to "roll your own", be aware that layout is critical, and
that component lead lengths have significant effects. This is why I
pointed you at the kits - all the layout problems have been "designed out"
for you.
My preference is for a VCO running at half the required output frequency,
followed by a diode doubler (needs no adjustment), a fixed high pass
filter, and then some stages of broadband amplification. I get to 5 Watts
with 4 transistors, and can get the same output power anywhere from 87.5
MHz to 1080 MHz /without/ /adjustment!/
There are two advantages of running the oscillator at half the output
frequency - there's no chance of a large PA stage coupling back into the
oscillator causing instability, and the oscillator is operating well inside
the range of cheap 74HC-series ICs for a phase locked loop.
My cheap PLL uses a 74HC4040, a 74HC4060 and a 4046, with a mostly passive
loop filter (has one transistor for bigger voltage swing) and using a
standard, cheap crystal (I use 4 MHz). The division ratios of the 4040 and
4060 are set using diodes, so frequency change involves cutting or
soldering diodes (a minor disadvantage in my view).
The VCO is a Hartley type, followed by a standard diode doubler - if you
match the diodes (ordinary 1N4148's work fine), you can suppress the
fundamental by > 40 dB and the high pass filter preceding the next stage
improves this by another 18 dB. The amplification stages are broadband and
biased towards class B - this reduces efficiency compared to critically
tuned class C stages, but eliminates alignment. The output of the last
stage passes through a classical half-wave filter to reduce harmonics to
acceptable levels.
This approach gives a 5 Watt exciter with < -90 dBc spurs, < --70 dBc
harmonics, stability to a few Hz and *no* alignment!
Bob
Bob
> I get to 5 Watts
> with 4 transistors, and can get the same output power anywhere from 87.5
> MHz to 1080 MHz /without/ /adjustment!/
Make that 108.0 MHz!
Bob
WSQT
an alternative to a spectrum analyzer, and it is actually easier than
aligning a receiver, because you already have your "signal generator!"
Here's how I align a "Generation 10" premix board like the ones in
both WSQT FM transmitters. "Generation 10" is of course my tenth
design for these, and works so well a TV can be watched on a channel in
which a spur(NOT a harmonic) would fall(channel 5 with one crystal, for
instance) while ten feet from the antenna at full power. Harmonic
suppression(channel 7 over 88 MHZ) is the job of the final and filter
of course, not the exciter.
You need a wavemeter/dip meter and/or a general coverage receiver
covering all your oscillator frequencies, the FM spectrum, and the
aircraft band. First, peak up both oscillators and verify they are
running on their intended frequencies. Next tune the mixer primary tank
for peak output at your final frequency, then the secondary tank tuning
and coupling. After than, touch up secondary and then primary tank
tuning.
Make SURE the mixer is NOT on a different mixing product, as there
will be the plus and minus mixes plus the harmonic mixes available!
The output from just one doubly-tuned, critically coupled tank
circuit will be clean enough to avoid confusion in aligning the amp
chain. Again, tune primary first, then secondary, then touch both of
them up to peak output.
ABOUT CONSTRUCTION:
These tuned circuits are simple-two coils side by side on the board
of 3 or 4 turns, about 1/4 inch long and 1/4 inch diameter. Use 3 turns
for a single ended stage and 4 for a push-pull stage(balanced mixer or
tripler), plus either a trimmer capacitor on each or fixed capacitors.
You tune with fixed caps by spreading or conpressing the turns of the
coil with a plastic tool after findign the approximately correct
capacitance. The latter method is used in a lot of commercially made
VHF devices, BTW. It is cheapest to make but hardest to tune a
prototype.
For a bipolar transistor feeding another bipolar transistor at low
levels, tap the secondary coil at the 1/3 mark, and connect that to
your next stage base. Most of the time, I found that having the
windings just barely touching one another gave critical(most gain
before selectivity suffers) coupling. Start with the coils 1/8 inch
apart, and check output closer togehter and farther apart(just bend the
leads).
This essentailly duplicates a receiver IF strip-but at your final
frequency, similar to an old TRF(tuned radio-frequency) receiver. While
it will NOT block spurs one, two, or three channels away like it does
in a receiver at 10.7 MHZ (or 455 KHZ for AM)IF, it WILL shut up all
those pesky spurs 3-8 MHZ away that otherwise devil a premix or
hetrodyne board.
If you want to use a Part 15(Ipod, etc) transmitter as a signal
source, run it by itself and put a receiver RIGHT NEXT to it, and take
note of EVERY PLACE you can hear it. Some places ar emade in the
transmitter, and some in the receiver. If you don't hear anything
within 5 MHZ of the carrier, you can filter it with this method.
High-Q stripline filters are another approach that can work for this,
even at high levels like between a dirty exciter and a PA. In one test,
a filter would allow a ham signal somewhere near 53 MHZ to pass with
less than 2 dB attenuation, but give something like 40 dB at 56 MHZ!
BTW-it you ever build a 2M ham rig using the hetrodyne layout, I
suggest double conversion to allow effective filtering with an
oscillator below 10 MHZ for stability. This is again essentially a
receiver in reverse and is engineered the same way.
Bob
> For aligning an excitrer dependant on critical filtration, there is
> an alternative to a spectrum analyzer, and it is actually easier than
> aligning a receiver, because you already have your "signal generator!"
>
> Here's how I align a "Generation 10" premix board like the ones in
> both WSQT FM transmitters. "Generation 10" is of course my tenth
> design for these, and works so well a TV can be watched on a channel in
> which a spur(NOT a harmonic) would fall(channel 5 with one crystal, for
> instance) while ten feet from the antenna at full power. Harmonic
> suppression(channel 7 over 88 MHZ) is the job of the final and filter
> of course, not the exciter.
Your approach is _entirely_ valid, but /not/ really suitable for a beginner!
_Harmonic_ _suppression_ is nothing (really) to do with the exciter stage,
but spurs certainly are - and the cheap "stereo transmitter" ICs
produce /really/ nasty output spectra. These things _are_ _not_ the way to
go!
It's actually easier to "tame" a PA stage if you have "clean" drive, so it's
usually worth applying some minimal harmonic filtering to the exciter stage
- I often used a quarter wavelength (multiplied by the velocity factor) of
coax between the exciter and the PA.
If you're using modern components, designed for use at VHF frequencies, the
broadband approach is very attractive. With my method, it's easy to get a
transmitter *completely* *free* of adjustments - even the VCO coil is a
fixed type! If you need frequency agility, my cheap PLL isn't the way to
go, but there are simple cheap "PIC and a PLL IC" approaches that give easy
frequency setting.
I fully understand your desire to push components to frequencies they were
not intended for, and your desire to make really cheap transmitters, but I
find that it's often a false economy. You'll spend lots of time adjusting
your critical alignment points, and lots more time experimenting with
methods to get the parts going in the first place. If you use components
designed for the job, you'll get working gear /much/ more quickly!
If you shop around on the 'net and are prepared to import, you can get some
great deals on RF components. I recently got some BLF177 FETs for <$12
each. This device gives > 170 Watts for slightly under 2 Watts of drive
(in a critically tuned design), so eliminates one whole stage in a
conventional transmitter. If you use it broadband, the gain is lower, but
1.5 Watts of drive will yield 120 Watts out *without* *adjustment!*
Bob
I wouldn't have your trust in wavemeters and general coverage receivers - a
spectrum analyser /really/ /is/ essential for a design as critical in
alignment as yours.