The TDA7088 is a simple Fm radio on a chip.
http://www.datasheetcatalog.org/datasheet/unisonic/TDA7088.pdf
Not the greatest,but fairly simple. Also,these (or a clone) are commonly used in those cheap FM pocket radios that you can get at the 'dollar store'. They have a Scan and Reset button.
I've bought a few just to tinker with,and made one with adjustable tuning,instead of scan tuning,just by adding a 10turn pot.
Also,These apparently spit out the full MPX signal,so with a decoder chip you can get stereo..maybe even RDS too.
http://www.datasheetcatalog.org/datasheet/unisonic/TDA7088.pdf
Not the greatest,but fairly simple. Also,these (or a clone) are commonly used in those cheap FM pocket radios that you can get at the 'dollar store'. They have a Scan and Reset button.
I've bought a few just to tinker with,and made one with adjustable tuning,instead of scan tuning,just by adding a 10turn pot.
Also,These apparently spit out the full MPX signal,so with a decoder chip you can get stereo..maybe even RDS too.
Sample and hold stereo decoders can run into problems with the sinc frequency response, which can give an HF rolloff. This will be worse than a simple NOS DAC because of a sampling rate of 38kHz instead of 44.1kHz.forr said:
Also, I suspect that they will be more affected by noise - which is the opposite of what is sometimes claimed for them?
The best place to find a textbook example of an FM receiver circuit is in an old electrical engineering textbook. There's actually only one way to built an FM radio, the block diagram is simple enough. Antenna, frequency tuning circuit/RF amplifier where the tuning capacitor is ganged to the oscillator frequency capacitor (the difference must always be 10.7 mhz), oscillator, mixer, IF amplifier, FM detector (Foster-Seeley discriminator) demultiplexer (optional for stereo FM) audio preamplifier/amplifier, speaker. There are an endless assortment of variants in execution and embelishments of each but that's the only strategy I know of. Advances came from among other things quartz oscillators, ceramic IF stage filters, putting all of the active elements (transistors/tubes) on an IC but it follows the same principles and functional block diagram.
Old "classic" designs were a PITA to align. You had to adjust everyting by hand with an RF generator, stereo generator, frequency meter, etc. You had to align everything, the coils in the oscillator, IF stages, and de-multiplexer. The goals were high sensitivity, optimal IF bandpass characteristics with good selectivity, and good channel separation with an absence of chirps and whistles from the de-mux. Except for those who like to tinker with old equipment, it's probably becoming a lost art. But that's nothing. Try aligning an old color TV set
BTW if you can't find one in a textbook, SAMS published schematics for thousands of different models. You can usually find them on ebay for a few dollars, or in some packrat teckies' basement.
There is always room for innovation, even in old technologies like FM broadcast. For example: using Double Balanced Mixer ring modulator at the front (no variable capacitors or varicaps), using programmable crystal oscillator, using linear-phase LC IF filters, using PWM "scaler" FM demodulator, using discrete or CMOS analog switches for MPX decoder, using vacuum tube audio stage, etc. etc. But who cares if today's broadcast quality and program choice is so poor?
I built an FM tuner using a DBM mixer, but still used varicaps for tuning. I doubt if you could make an imageless mixer good enough for FM band use, and it would be unnecessarily complicated. Possibly worse performance than using varicaps.
I did think about using CMOS switches for the stereo decoder, but eventually took the easy route and used a Hitachi chip. As you say, broadcast quality is now so poor that a good FM tuner can be a mixed blessing as you can hear all the mistakes - even the BBC now routinely does bad sound.
I did think about using CMOS switches for the stereo decoder, but eventually took the easy route and used a Hitachi chip. As you say, broadcast quality is now so poor that a good FM tuner can be a mixed blessing as you can hear all the mistakes - even the BBC now routinely does bad sound.
Image rejection is basically a combination of filter rejection and good separation between the local oscillator and received band..
Bear in mind that a lot of the more "modern" detectors are not developed with sound quality as the main focus, rather than avoiding "man intervention" in the form of manual tuning. One of the best FM detectors are still the double slope detector, but unfortunately a design that requires a lot of tweeking and tuning to work optimally. A truly modern detector design would probably use DSP for the main functions. But what is all this good for, when most broadcast services are going for *"#**@ DAB !
This is a very comprehensive design, and a very good but short description of the design culprits. Unfortunately some of the components may not be found these days.
http://mobiledevdesign.com/images/archive/1000Ryder26.pdf
Bear in mind that a lot of the more "modern" detectors are not developed with sound quality as the main focus, rather than avoiding "man intervention" in the form of manual tuning. One of the best FM detectors are still the double slope detector, but unfortunately a design that requires a lot of tweeking and tuning to work optimally. A truly modern detector design would probably use DSP for the main functions. But what is all this good for, when most broadcast services are going for *"#**@ DAB !
This is a very comprehensive design, and a very good but short description of the design culprits. Unfortunately some of the components may not be found these days.
http://mobiledevdesign.com/images/archive/1000Ryder26.pdf
Marantz 10B designed in the early 1960s is still considered a fine tuner even by today's standards. The concepts may be antiquated but the results are still pleasing to many. The main contribution to satisfactory FM reception is a good antenna. A mediocre tuner with an excellent antenna signal will beat a top notch tuner with a poor antenna signal every time. The more modern implimentations have eliminated the need for tedious and repeated alignment.
The best tuner in the World will sound terrible if the RF signal is weak and full of multipath interference.
Badly designed tuners can have too much gain before the mixer and IF filters. These will still work better with a good antenna to select the wanted station and then an attenuator to avoid overload.
Badly designed tuners can have too much gain before the mixer and IF filters. These will still work better with a good antenna to select the wanted station and then an attenuator to avoid overload.
Here in the UK all our stations tend to come from the same transmitter mast, so the antenna cannot select a wanted station. Maybe elsewhere a very directional antenna mounted on a rotator can pick stations coming in from different directions.
A bad tuner (noisy local oscillator, poor alignment, too narrow IF, non-linear detector) will still sound bad whatever signal you give it.
A bad tuner (noisy local oscillator, poor alignment, too narrow IF, non-linear detector) will still sound bad whatever signal you give it.
That was EC8010 thought, see posts #15 and #17 :
http://www.diyaudio.com/forums/analogue-source/95310-hybrid-fm-multiplex-decoder.html#post1123888
His original idea was a switching decoder, but then the conversation shifted to sample-and-hold. It is possible that some people thought they were the same thing - I'm not sure. Both decoders have problems, but different problems.
Sample-and-hold suffers from a sinc frequency response, which is phaseless. Unfortunately this means that any attempt to correct it will introduce phase errors. On balance, probably not too bad. I saw a circuit somewhere (probably Wireless World or HiFi News) which used this method without any correction. I suspect the author did not realise what was happening, and may not have realised that the 'smooth' sound he got was due to HF rolloff.
Switching decoders give incomplete separation, because a pi/4 creeps into the maths, so a bit of cross-mixing is needed to cancel out the error. Decoder chips will already include this, but some DIYers might not realise that it is needed. You get 19.4dB separation without the extra cross-mixing, which will sound reasonable but measure poorly.
Interestingly, you could make a half-and-half decoder which gates for 13.16us instead of 26.32us but still at 38kHz rate. This would give a little bit of both problems, but perhaps not enough of either to worry about?
Sample-and-hold suffers from a sinc frequency response, which is phaseless. Unfortunately this means that any attempt to correct it will introduce phase errors. On balance, probably not too bad. I saw a circuit somewhere (probably Wireless World or HiFi News) which used this method without any correction. I suspect the author did not realise what was happening, and may not have realised that the 'smooth' sound he got was due to HF rolloff.
Switching decoders give incomplete separation, because a pi/4 creeps into the maths, so a bit of cross-mixing is needed to cancel out the error. Decoder chips will already include this, but some DIYers might not realise that it is needed. You get 19.4dB separation without the extra cross-mixing, which will sound reasonable but measure poorly.
Interestingly, you could make a half-and-half decoder which gates for 13.16us instead of 26.32us but still at 38kHz rate. This would give a little bit of both problems, but perhaps not enough of either to worry about?
If you want to get into the nitty-gritty of fixing tuners, I hear good things about ""The Theory Of Servicing AM, FM, And FM Stereo Receivers" which is discussed in the thread by the same name:
http://www.audiokarma.org/forums/sho...d.php?t=495126
You can DIY a very good chip-based FM decoder and use it with any number of tuners including tube units:
fm-mpx : FM Tuner MPX Kits
I have one and it works well and sounds good!
http://www.audiokarma.org/forums/sho...d.php?t=495126
You can DIY a very good chip-based FM decoder and use it with any number of tuners including tube units:
fm-mpx : FM Tuner MPX Kits
I have one and it works well and sounds good!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.