there are two version of this rds with slight difference, the one i made had a 10 ohms resistor at output i have change it to 1k and voila i have set the level being very delicate between 2 and 3 khz deviation , i have not yet put any filtering , lazy i am .i prefer adding an external extra filtering apart alone .What is the best level rds or standard recommend , btw if you change the exciter board , the level must be adjusted , even if the exciter is nearly set to 0db

RDS level?
Recommendation ITU-R BS.643-3:
"Sub-carrier level: the recommended nominal deviation of the main FM carrier due to the modulated sub-carrier is ±2 kHz. However in practice it can be as low as ±1.2 kHz; many EBU broadcasters with wide dynamic range (e.g. classical music) services prefer this lower value to ensure best signal-to-noise performance. The decoder should, however, be designed to work with sub-carrier levels corresponding to between ± 1 kHz and ± 7.5 kHz deviation."

When I set the RDS I set the level to 6% (100% = 75KHz). Although car radio, HiFi, and portable radio can detect RDS with a level as low as 3%, smartphones has the RDS threshold 5-6%.

thank you mikroman for the valuable infos .
Does RDS in mono eat a lot of rf power , or this have not any issue with the broadcasting , my transmitter is about a `100w ,transmitting in mono should be , not ready yet , old fashion the best i can achieve at the moment .
i use a 300 mw of RDVV tuned to near 0db with an input of 1000 cycles tone ,then the same tone is re-entred to the rds and achieving the calibration to near 0db all seen on the p175 analyser
then i can get the rds level set to 2,5 khz but i have a doubt of 1 resistor marked 3k where i have seen it about 30k and i have fitt my bad eyes 33k hhh .

RDS is on a diet so you won't feel any difference in range

Ok so i done some quick tests!

The first is just a baseline reference to indicate if any spurs are present when the RDS is off.

Then we have a capture of each with different span range 0-98KHZ 0-195KHZ And 0-390KHZ

As expected the harmonics are present, But also spurs across the band can be seen!

Although these spurs are less significant? They could contribute to audio noise?

These are also unloaded values so may differ when the unit is put into operation?

Anyways, Take a look and see what you think?

Sinus i see no diff between the last 3 , there is indeed harmonics , maybe an external lc tuned filter will do the jobes , but i have achieved a ev of 2-3 khz wich is good as mikroman advice , and i will try what Albert also sugest ,how i can try a resonnat circuit with my little dso oscillo

Sinus for curiosity what software do that ? does it need a hardware or just an audio probe to soundcard with line in

To be honest Radium, It sounds like you have your system running pretty good!

You could filter and filter a bit more? But i personally think it would be a waste of time and money, Especially if the source itself (PIC) is noisey to start with!

All of these RDS units are designed for one purpose only! That is to be cheap and simple!

High precision and performance would be hard to find at this level!

Oh The Spectrum Analyser is a Picoscope2205 which is a PC based Oscilloscope.

Thank you Sinus as always , i think i still have to add extra filtering , but i am not sure how i can see in practice if that circuit diy LC resonate at 57 Khz i only have a little dso138 oscilloscope any help .I am trying to figure that out ,but some doubts .
another thing , i have made a little error replacing a resistor of 30k by a 33k so i will correct it but i still have a little doubt with another resistor value if you can give me a hand i will post the error resistor in question .

Hello Radium

I have never experimented with that version before? If you are thinking of adding an LC filter? You can easily calculate its resonance manually or via an online calculator?

However remember that you could compromise the system impedance by doing so!

The 'Correct' Resistor you have highlighted seems to be part of a potential divider? I guess to set the gain for the following stage?

The other resistor highlighted seems to be part of a 'T' filter? Again the impedance of the stages it lies between will determine values you need for the filter to be effective!

If you dont fancy doing the maths? You could use temporary potentiometers to trim to your preference?

Regards to a DSO? Most have a spectrum analyser capable of sampling up to 100Khz or more!

thank you sinus , have a calm end of week , keep safe.but what i ask is i saw many picture that this 3l resistor is replaced by a 30k , so is it wrong on the schematic

radium98 wrote: ↑Sun Apr 18, 2021 7:27 am thank you sinus , have a calm end of week , keep safe.but what i ask is i saw many picture that this 3l resistor is replaced by a 30k , so is it wrong on the schematic

Well this would be a guess?

A 3K (Or 3.3K) is indicated on both the schematic and layout diagrams which would not heavily load the TL072 output? Neither will the LM386 input present any significant load or the 10K potentiometer?

30K sounds a bit high and i would go with what is on the schematic!

Gain will be affected either way? So any kind of filtering will reduce the RDS signal aswell as the noise!

thanks Sinus

Hello .Is there a way to synchronize the 19khz that come from an external ST encoder with the 57khz of this Mon rds with the last schematic posted Up .using a little extra pcb that have the famous cd4046 and how should be the connection , if you can give me a drawing or a hand .or even using maybe the lm567 ,but i prefer 4046 or any other phase comparator
thanks

I synchronised my RDS by deriving the 4332kHz signal from the same crystal source that was used for the 19 & 38 kHz signals for the coder. I can put the circuit up if you want - it's just a bit of CMOS logic.