Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

Which stereo encoder?

Status
Not open for further replies.
The first circuit tries to use old 5V TTL with 12V Cmos but they do not go together. It also uses very old 741 opamps which are horrible for this high gain at high audio frequencies circuit.

It has a lowpass filters at the inputs that cut 8300Hz by -3dB and higher audio frequencies are cut more. FM is supposed to be flat to 15kHz but the old 741 opamps do not go that high anyway.
It has 33k parallel with 1.2nF for pre-emphasis which is good but since it cuts high frequencies then the pre-emphasis will not do much.

The second circuit is missing pre-emphasis so an FM radio sound will be missing high audio frequencies and will sound awful like an AM radio. it is also missing audio lowpass filters to reduce audio frequencies near 19kHz so they will beat with the 19kHz stereo pilot tone and cause whistles.

A stereo encoder needs to feed an FM transmitter with an audio response flat from 30Hz to 57kHz. Do you have one?
 

So the first one is a no go. I already have a filter design and preemph so I am not worried about that. So, the second one will work? I do not want to use an NJM2035 cause I cannot get one to work without a lot of noise on the receiver end.
 

If this is not for professional use, I would go with the first design but with better op-amps and a more consistent pilot tone generator. As pointed out, it is rather silly to use a mix of TTL and CMOS ICs although they do use a crude level converter between them. A crystal controlled CMOS oscillator will work just as well with fewer components. Personally, I would start with a higher crystal frequency and divide it down more, 76KHz crystals might be hard to find, higher frequency ones will be cheaper and more easily sourced.

It might be worth your while looking at some of the dedicated stereo encoder ICs, their audio quality isn't fantastic but may surpass the designs you posted.

Brian.
 

Simulation shows that input audio amplifier will work OK. Pre-emphasis cap 1.5nF will give 50us (40us now). Simulation with TL071 shows useless gain/frequency characteristics. It seems that amplifier design perfectly suits to LM741 OPA.
Between 74ls390 and 4027 is level translator which enables triggering CMOS FF from 5V TTL.

MPX filter has no data for L1 and L2 ( about 45mH?) and transmitting in neighbour FM channels will happen.

Your transmitter must have linear ferequency characteristic up to 57kH within 0.1dB if you want channel separation to be 40dB.

On the same site you have PCB layout etc.


betwix suggestion is good. The same site https://electronics-diy.com/store.php has designs/kits of complete stereo transmitter.
 
Last edited:

Thanks for the help, I might have to try both circuits just to have fun and see what goes on. Do not have Oscope ( have bills to pay so...) so I just do everything by experimentaion. I do actaully have 75Khz crystals. This is not professional by no means at all. I have looked into the NJM2035 and the BA1404 and both operated pretty well but the NJM and BA1404 had to much pilot tone as it sounded like. Again no Oscope so cannot be sure 100%. In the datasheet of the NJM there are filters ( Band pass and low pass) but cannot bring stereo from the filtered design, I know that there are adjustments to be made but cannot seem to make them. And I know about the 15Khz filters on the inputs so that is not a problem as I have a working design ( according to simulation ). about 45mH you say? I have 20mH as my highest. Lets see If I can make that work. I am guessing it is a low pass filter for the 19 Khz pilot tone?
 

I do actaully have 75Khz crystals.
Click to expand...

Is that a typo? The crystal has to be 76KHz ( 4 * pilot tone) or it won't work.

I'm not sure what you mean by too much pilot tone, the level isn't critical, too low will make your decoder drop to mono, too high will cause high frequency beats with upper audio frequencies but most receivers are tolerant of quite wide ranges.

Brian.
 

I simulated the first circuit's input lowpass filters and pre-emphasis using a better opamp than the lousy old 741:

1) There is way too much audio level near 19kHz because the lowpass filters have responses that are too gradual. This causes whistles in the output of the encoder.
2) There is not enough boost from the pre-emphasis at high audio frequencies because the lowpass filters have responses that are too gradual.

It has a peak at about 28kHz because the pre-emphasis keeps boosting at ultrasonic frequencies, then the opamp's high frequency response drops at such high gain together with the lowpass filters.
Its main problems are that the value of R4 (560 ohms) is much too low and that its very gradual lowpass filters cut high audio frequencies as much as they cut frequencies near 19kHz.
 

Attachments

  • FM stereo encoder response.png
    FM stereo encoder response.png
    25.7 KB · Views: 211

Betwixt-75Khz was a typo. Meant 76 yes. I'm not sure what I really meant, or just not sure what of how I was trying to phrase "to much pilot tone". Just at the receiver end I get a lot of whistling on the receiver end. And when the band pass filter is connected on the pilot pin the stereo drops out to mono. Here's the datasheet to the njm2035

http://pdf.datasheetcatalog.com/datasheet/newjapanradio/de05003.pdf
Audioguru- what opa did you use? And I wasn't going to use the 741 cause I do not have any. So, according to your drawing, that is not a great filter. I'm leaning on using the first drawing I posted if I can find the inductor values. The second drawing link, out of curiosity. Would it work?
 

The NJM2035 and BA1404 ICs are too simple to make good FM stereo. Rohm stopped making the BA1404 years ago because of complaints about it then made the much better and more modern BH1415 series of FM stereo transmitters.

High audio frequencies (pre-emphasis has boosted them) beat with the 19kHz stereo pilot tone which causes whistles. So the circuit should have a very steep lowpass filter (called a brick wall filter) that cuts frequencies above 15kHz. The inductors in the first drawing smooth the 38kHz switched audio channels so that their many harmonics do not use up output level and cause adjacent channel interference.

I have used TL071 single, TL072 dual and TL074 quad low noise audio opamps for audio. They perform well up to 100KHz and have very low distortion. Their inputs are Jfets so their input bias current is nothing and their input resistance is infinity.

The second drawing modulates the L-R at 38kHz, filters it then adds it to the L+R. Maybe it should filter the L-R before it modulates it.
The first drawing filters the audio channels not enough, simply alternates them at 38kHz then filters the output.
 

Yeah, I do not Like the njm or the ba1404 anyways. I cannot find anything online as it seems everything has problems. If you have a good circuit I could use that would be great or I'll have to figure out how to improve the 2 linked circuits.
 

A magazine in Australia posted The Micromitter which was an FM transmitter project 10 years ago that used a BH1417 stereo encoder IC. A kit was available and it worked fairly well. The IC might not be available anymore today.
Today I noticed an article about using modern DSP in a stereo encoder and it said something nasty about the BH1417 IC.

I looked for FM Stereo Encoder Circuit in Google Images and found some good circuits.
 


I think both circuits you found will sound bad.
Harry's circuit has poor pre-emphasis that barely boosts high audio frequencies. The boost at 15kHz in Europe should be +14dB but it is only +0.7dB because the value of C2 is too low and the value of R1 is too high. There are no filters.
 

Yes, I threw Harry's circuit together and it did sound terrible. Decided not to go with that one. The other circuit I do not have crystal for so did not try but if you say it is bad then it is bad. These 2 circuits keep showing up in google images and people are tearing them up.

https://www.electroschematics.com/315/stereo-encoder/
https://www.electroschematics.com/325/stereo-encoder-high-separation/

I can tell both circuits do look terrible just by the fact that they lack proper filtering.
Seems people just post whatever they want and call it a working circuit.
 

Welcome to the Internet 8-O
Now you know why we are so critical of many of the designs posted on it. On that second 'improved' encoder I wonder what the LC filter really does. Obviously it is there to peak the pilot tone but the values around it make it very ineffective.

Brian.
 

Please do not post any more hopeless circuits.
One says it has 70% channel separation. What is that? Maybe if only one channel is supposed to have sound then the "quiet" channel has a level of 70% of the "loud" channel which is a separation of only 3dB?
You would barely notice it.

An FM encoder should have separation of 40dB which is levels that are 100:1.
 

BA1404 is best solution for backyard FM transmission where low power FM transmitters are legal.
 

Borber said:
BA1404 is best solution for backyard FM transmission where low power FM transmitters are legal.
Click to expand...
But the BA1404 is obsolete. It has not been made for about 10 years when Rohm replaced it with the much better BH1415 series of frequency-synthesized FM stereo transmitters.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top