Audio Op Amp - Curing RFI Feedback Loop????

Hi,

Ok so here's a problem that's been bothering me for quite some time and I think after much hair pulling and extensive testing I've located the actual cause of the problem but I am unable to find a suitable solution at this time and would like some help please.

The Problem:-

I am using a DFPlayer Mini MP3 Player in close proximity to RF which the MP3 player is picking up and reacting to. These MP3 modules are very simple things. They use a dedicated YX5200 MP3 decoder IC with on board DAC and Serial Comms and the audio from that is fed into an 8002A Audio Op Amp.

The problematic wire seems to be the Op Amp's Negative Speaker Wire Output which also feeds back into the negative input on the op amp via a 68K resistor. No doubt this is feedback for the Amplifier Gain.

If RF gets anywhere near this Speaker Negative wire, a buzzing / humming starts to appear and gradually increases in strength and begins to loop over and over. When the RF is removed, this buzzing carries on looping round and round like an echo but it starts to decay.

This suggests to me that the negative output is picking up the RF and looping it round and round through the feedback back into the Op Amp input.

What I've tried to Resolve the Problem Myself:-

First of all, I should point out that when I built my project, the MP3 gave me a great deal of difficulty to begin with. Until now I haven't understood where the problem actually lay so please don't mock my silly ideas. I'm still learning lol.

So when I first built my project, the Speaker Negative output from the Op Amp has always given me trouble. No matter where I connected it or how I connected it, I would get dreadful buzzing so I left it free floating and just used the Speaker Positive output from the Op Amp. And that got me some audio out but that came with an equal level of buzzing / humming noise.

Next I decided to try shielding so I put my project in a metal enclosure. That cured a little of the noise but plenty still remained. So I added layers of tinfoil between the enclosure lid and the PCB which dropped the noise considerably but some still remained.

Then I decided to us a DC blocking capacitor in series with the Speaker Positive which cleared up all but a distinct audible buzz at a certain frequency which I couldn't get rid of at all and this is when I decided to investigate the problem in more depth.

As turns out, ignoring that Speaker Negative wire was a BIG mistake and all it does is pick up RF and feed it back through to the amplified audio signal over and over and over.

So stupid me now has a problem of my own making of sorts when I should have just dealt with the issue on day one and found a way of dealing with the output from the Op Amp properly!

Lesson learned there....

Further Research Into the Problem and Possible Solutions:-

Now that I understand what the problem actually is, I can stop stabbing away in the dark and hoping I get lucky and actually do something about it.

Having done some reading, it seems I need to use RC filters at the input of the Op Amp to keep the RF at bay but I must confess, I've no idea where to start.

Ideally I'd like to modify the MP3 modules I'm using but due to their small size, that doesn't look like it's going to be a simple task and will require some seriously steady handed soldering!

The other option is to get rid of the MP3 modules and source some MP3 decoder chips and build my own MP3 circuit with filtered Audio Op Amp. I'm happy to go that route but I have no idea how to go about filtering the RF from the Op Amp.

I've read some about it and it says to use RC filters to keep the RF out of the Op Amp's input but I've no idea where to begin if I'm honest. It's all going round in my head and I'm trying to make sense of what I need to do.

All I need to do is get a good audio signal from the MP3 file into the AF input on the transmitter without the RF affecting it.

I did find this article on filtering Op Amp input which looked promising until I noticed that the Images that go along with it aren't in the article or on the server anymore :frowning:

http://www.electronicdesign.com/analog/put-whammy-rf-noise-emi-without-hurting-performance

Any help I can get on this I would greatly appreciate!!!!

Also another issue I have is that I can't find any MP3 decoder ICs to buy at a reasonable price anywhere. The YX5200 seems to be obsolete. The next version seems to be the YX5300 but again, you can't buy those IC's anywhere either. I notice at one point Atmega did their own MP3 decoder IC but that's also obsolete.

Perhaps my search powers are beginning to fail me but I can't find any cheap MP3 ICs anywhere.
All google brings up are ones on eBay for silly money, ones that are obsolete or it brings up complete MP3 modules. What's going on? Is MP3 obsolete or is there some new fangled technology I'm not seeing?

So my only choice right now seems to be to buy MP3 Modules and butcher the IC's from them.

All help greatly received!!!

Thanks!

OK if you need help you are going to have to be a lot more specific. What op-amp are we talking about? Most do not have a positive and negitave output. What is the schematic of this amplifier?
How is it powered? How is the RF module powered?
Is their any power supply decoupling on you system? How is it physically layer out ( photo please ).

In general What you describe as a slowly decaying sound does not seem to be what you think it is because the time constants are just too long for that.

Hi,

Can you post a picture of your project so we can see your component layout?

Can you please post a copy of your circuit, in CAD or a picture of a hand drawn circuit in jpg, png?

Thanks.. Tom... :slight_smile:

Hello and thank you both very much for taking the time and trouble to respond.

UPDATE:-

Since writing my original post, I have done further testing and have stripped my project right back until I was left with the transmitter module and power supply (3A DC-DC Buck Converter @ 4V).

I then took the transmitter module to one side, powered it up from the Buck Converter and with no audio connected, I put out a blank carrier and that annoying 200 - 250Hz buzzing was present and still as strong (showing around -50dB on the SDR I was using to monitor).

I also noticed lots of similar peaks spaced equally apart moving up in frequency which decreased in strength the higher up in frequency they appeared. It looked like and probably is some sort of harmonic interference....?

Further Troubleshooting Steps & Results:-

  • I have been using a rubber duck style antenna since day one but to see if that was anything to do with the problem such as the RF being too close, I connected the transmitter to an externally mounted antenna but the result was exactly the same.
  • I next considered that the Buck Converters might be being affected somehow so I powered the module up from a good known working adjustable bench power supply capable of delivering up to 35V and 5A. Interestingly, when I put the transmitter into transmit, it transmitted briefly (1/2 - 1 sec and then cut out) - So I'm not sure what to make of that other than the RF got into the power cables and overloaded the Transmitter Module.

Tests Yet to be Carried Out:-

So now I've narrowed things down, this is pretty much all I can think of that I can try now.
Any other test suggestions would be welcomed because I'm rapidly running out of ideas now.
I have tried two modules so far and both do the same thing so unless I've damaged them both in the same way somehow, the problem isn't with the transmitter module itself but most likely my design(s) implementing it.

These are the things I've got left to try...

  • To rule out all these converters and power supplies, I am going to source a 3.7V battery pack and see what happens when I power it directly without any other electronics in the equation.
  • There is no filtering between the Antenna Output on the Module and the SMA socket. With there not being too much information available on these transmitter modules, I only had the design of others to reference. The datasheets are not worth mentioning they are so poor. Many of the designs I looked at which use these Transmitters do not have any Low Pass Filtering on the RF Out so I didn't bother either. I didn't see the point because I wasn't getting any issues until now. But in light of this, I've ordered a couple LCFN-160+ SMD Low Pass Filter Packages to try and see if that stops it

Transmitter Breakout Board Layout & Schematic + Power Supply DC-DC Converter:-

So this is what I'm working with and where the problem lies - somewhere.

Transmitter Module Breakout Schematic

Transmitter Module Breakout Board Layout

The board is made from double sided FR4 with top and bottom ground planes and some via stitching tying the top to the bottom. Any unused header pins that are not connected to the transmitter module have also been tied to the ground planes to try and form an RF sheild - of sorts.

No SMA edge mount connector is shown in the diagrams because it was a last minute addition. Originally I was using a very short fly lead soldered directly to the module with an SMA on the other end but I discovered later that I could just get an SMA connector to fit snugly without shorting anything.

SUN-D 3A Dc-Dc Buck Converter

And finally, this is the DC-DC buck converter I am using. I power it from a 12V SLA battery and step it down to 3.7 - 4V for the transmitter and other modules. It uses the MP1584EN which was the lowest consumption converter I could find. My project is battery operated so low power is a must and these converters draw very little with my project.

Thanks again!

Hi,
Can you post a picture of your project so we can see your component layout?

Can you please post a copy of your circuit, in CAD or a picture of a hand drawn circuit in jpg, png?

We need to know how you have connected EVERYTHING connected in a diagram including power supply.

Component layout and how you wire your power supply to those modules is very important.

What frequency is your Tx working at and do you have it transmitting into a TUNED aerial?

Thanks.. Tom.. :slight_smile:

What RF transmitter are we talking about ? What power and frequency ?

How long are the speaker wires? Have you tried a ferrite toroid over the speaker leads and
opamp input wires?
Is the 8002A over a groundplane?

Show us the circuit please....

Hi,

As you do your testing and thinking about the signals getting into the audio section, you should know the audio amplifier is not picking up and amplifying the signal. The RF signal is getting into the circuit and being detected by a / some transistor or diodes. So traditional amplifier solutions will not work, the RF must be stopped before getting to the amplifier. This might have been obvious but I thought it worth saying.

Typical RF tools are:

  1. 100 or 1000 pf caps to a ground plane (on any wire going into the audio section)
  2. Series inductor (toroids) or ferrite on the amplifier wires.
  3. Physical shield (can, groundplane etc)

Not knowing your layout, it might be an interesting to short the audio input to ground inside the module and see if you still get you interference.

Hi,

Please re-read my last reply (Post #3).

I have unplugged the transmitter from the rest of my circuit and have connected / powered it up separately and tested it on its own and the buzzing is still there. So the transmitter is interfering with itself - somehow. So I have established it has nothing to do with the rest of my circuit, including the audio Amp, which is all on a separate main PCB board and has been ruled out through process of elimination.

I have posted the schematic and layout for the transmitter module break out board (which I designed and built) but I cannot post a comprehensive schematic of the actual transmitter module itself because I can't find one. All I had to go on was the information I could piece together from what other people had found out through trial and error.

The transmitter is made by a company called Dorji and the Model is DRA818V. It is a pre-built, self contained VHF FM Transmitter module based on the RDA1846 FM Transmitter IC and is more or less a cut down version of a Baofeng Mobile Radio in a small module package.

This is a typical connecting diagram for the DRA818:-

The only real difference between the breakout board I made and that typical connecting diagram is that my breakout board doesn't have a LPF (Low Pass Filter) in series with the RF OUT (Pin 12).

And the reason I never included a LPF in my design was because when I was prototyping on breadboard in the early days of my design, I never had or certainly never noticed this buzzing so made the assumption that everything was good. But since I built the little break out boards for the transmitter module, the problem (buzz) has become extremely bad. Most likely due to the proximity of the antenna to the rest of the circuit without adequate filtering.

Researching in the Early Days:-

When I was researching these transmitter modules in the early days, I found lots of designs for Arduino APRS trackers using these modules and the majority were not using any kind of filtering at the RF Output so that combined with not having any issues while prototyping on breadboard seemed to validate my lack of need for any filtering - until now.

Since having my problem with this transmitter buzz, I've found other break out boards which have used vias stitched all round the perimeter of the RF PCB Trace to form an electrical RF screen, which is another thing I didn't do. I had thought that keeping the trace / wire on my breakout board short and the connection as close as possible to the RF OUT pin on the transmitter would prevent any RF leakage.

To Answer Some Questions:-

Yes, the antennas I am using are tuned. They are commercially built, dedicated Amateur Radio Antennas. The main antenna I have been testing with is a "rubber duck" style walkie talkie antenna for 2m/70cm (VHF/UHF) but I have also tested on a dedicated, externally mounted, mono band vertical VHF antenna with the same results.

The frequency range I have the transmitter working at is 144Mhz to 146Mhz (2M Amateur Band) - and before anyone points it out, yes I do hold a valid and current Amateur Radio License in my country so I'm not wildly and randomly transmitting where I shouldn't be :slight_smile: lol

What the Project is About:-

For those of you that are wondering / interested in what I'm doing / trying to achieve - well, primarily it's a project of learning and boy have I learned along the way!

Each new project I think up, I like to be more challenging than the last so I am forced to learn new stuff while putting it into practice. In at the deep end, sink or swim as it were. Usually I try and build on projects using components and modules I've gained experience from previously and then ramp it up a notch.

In the case of these transmitter modules, I've gone from previously creating a simple Fox Hunt Transmitter which worked without issue, to trying to build a small, portable VHF Repeater which is being quite problematic due to its complexity.

But on the flip side of all the issues I've had getting this far, I've learned so much! I know more about RC, LC, LR Filters than I did before I started and I've never so much as seen an Op Amp before I started this so I know far more about them than I did previously.

I am hoping to go for my full license next year but I'm not a fan of just read and repeat learning. I like to get hands on, see what things do, ask why and find out why by putting it in to practice. I find I learn far more when I'm building stuff and troubleshooting it than I do from just reading the course materials.

And now I'm so tantalizingly close to finishing this part of my project if I can just cure that annoying buzz.....

My next step is to de-solder the transmitter module from the breakout board and fit it on to a new breakout board which has a LPF in series with the RF Out and see if that helps.

So...in short.... are you saying that you have an RF receiver that receives encoded mp3 info.... and the decoded mp3 info goes to your mp3 player module......and some kind of noise is heard when your receiver module is in close proximity to the transmitter side of things?

Yeah... maybe as some people mentioned already.... try ferrite bead of rf chokes..... and also ensure good grounding of two sided boards. And ground tracks thick as possible.

Decoupling capacitors on ics are beneficial.

Hi,
Sorry..

This is a typical connecting diagram for the DRA818:-

Doesn't show YOUR COMPLETE ciruit.

Can you please post a copy of your circuit, in CAD or a picture of a hand drawn circuit in jpg, png?

Tom... :slight_smile:
(P.S. I said Please..)

I'm worried by the antenna output going on the same connector as everything else. The breakout should act as an impedance-matched launcher to get out of those module pins and into a proper coax connector.

Cut that trace and solder the connector over the top of the module?

Hi,

In the case of these transmitter modules, I've gone from previously creating a simple Fox Hunt Transmitter which worked without issue, to trying to build a small, portable VHF Repeater which is being quite problematic due to its complexity

  • Is this a full duplex repeater?
  • What are your Tx and Rx frequencies.
  • If it is in the 2m band then the standard Tx/Rx separation is 600kHz, what are you using?
  • How do you aim to prevent desensitization?
  • Have you got any bandwidth filtering on the AUDIO?

Tom... :slight_smile:

Thanks for all the input / advice. I have now solved the problem.

I switched from using the DRA818V Transmitter Module to using the SR_FRS_1WV Transmitter Module and redesigned my break out board around it.

I also included a Low Pass Filter (Mini Circuits LCFN-160+) between the Antenna Pin on the module and the SMA Edge Connector and then finally, I used VIA stitching around the outside of the antenna trace on the PCB from the Module's Antenna Pin all the way to the SMA connector.

I couldn't tell you which of those steps was the actual cure. Maybe it was one thing or maybe it was everything combined but it all works now. I also no longer need to exclusively house my project in a metal enclosure to combat RFI. With this new module and setup, I can transmit without the project being in any box and I get no interference at all - much to my amazement and delight I might add!

Also for the benefit of anyone experimenting around with these transmitter modules, I can highly recommend the SR_FRS_1WV over the DRA818V. The SR_FRS module is half the price of the DRA818V and in my personal opinion, the SR_FRS is much more user friendly and the audio quality seems to be far superior as well.

Thanks again!

Hi,
Can you answer the questions in post #11 please.
You have got the Tx what about the Rx to complete the repeater?

Tom.. :slight_smile:

TomGeorge:
Hi,
Can you answer the questions in post #11 please.
You have got the Tx what about the Rx to complete the repeater?

Tom.. :slight_smile:

Hi, sorry - I figured because I'd solved the problem, no one would have any further interest.

You'll have to bare with me because I don't work to convention and am actually prototyping in reverse. I don't have any system analysis or design docs, no pseudo code, no Venn diagrams or anything like that. I just have an end result in mind and work towards it, dealing with any problems I encounter along the way and adapting where necessary.

I am all more or less self taught so I have a workflow that baffles most people LOL.

So at the moment, I've just been working on the output stage in very simple form and then building on that.
Mainly to ensure everything worked together as I had hoped before getting too carried away - if that makes sense?

So at this point in time, I only have a basic TX board prototype which comprises of an Atmega328P-PU, an MP3 Module, a Real Time Clock and a transmitter module. And all I've got it doing so far is running a simple test program which transmits a morse ident every 15 minutes which is played from the MP3 module.

Because I anticipate the code becoming rather large and complex in the end, I wanted to save space and resources and struck upon the idea of using these stand alone MP3 modules to handle the morse and voice idents with pre-recorded MP3 files which the board transmits at set intervals. So rather than having to alter code to change the morse ident every time, I just change the MP3 file(s) to suit.

The other reason for opting to use the MP3 module was to make the project more user friendly without people needing to alter too much code. So they can just make their own mp3 recordings, name them accordingly and insert the SD card and they're good to go. If I ever get anywhere with this project, I hope to release it open source - once I've ironed out the basics and proved the concept.

So that's where I'm at at the minute with it. And now I've managed to rid the interference problem, I can carry on moving forwards.

Believe it or not but On and off, this project has been almost 2 years in R&D...
I'm only a hobbyist and I've had to learn everything along the and I've made a number of alterations / improvements along the way as I've made new discoveries.
I'm now light years away from where I started believe it or not. I originally started out chemically etching boards which were rather large and used through hole components and using rather large hand held baofeng walkie talkie radios controlled by relays and now I'm milling my own PCBs using surface mount componants and MOSFETS for power management where before I couldn't tell you anything about a MOSFET lol. And I've also managed to get the standby power consumption down to around a milliamp or two along the way which was one of my primary goals. It's all taken me a lot of time but I'm getting there lol

So in short, to answer your original question lol, I've not got as far as the RX side of things yet, I'm sort of going with the flow and letting one design influence the other, if that makes sense?
But now I know the TX board can work and nothing interferes with anything else I can start working towards developing that side of the project.

Thanks again!

Hi,
Okay, so can you answer the questions in post #11?

You need to establish these to build your repeater.

It will not involve masses of code, just a timer and MP3 player in code.

Tom... :slight_smile:

TomGeorge:
Hi,
Okay, so can you answer the questions in post #11?

You need to establish these to build your repeater.

It will not involve masses of code, just a timer and MP3 player in code.

Tom... :slight_smile:

TomGeorge:

  • Is this a full duplex repeater?
  • What are your Tx and Rx frequencies.
  • If it is in the 2m band then the standard Tx/Rx separation is 600kHz, what are you using?
  • How do you aim to prevent desensitization?
  • Have you got any bandwidth filtering on the AUDIO?

Hi Tom, sorry, I missed those questions but will try to answer them here.

  • Half Duplex for now - only one person may transmit at a time while others listen
  • For testing purposes the Input Freq is 145.025 and the output is 145.625Mhz - locally to me there's no repeater activity there.
  • 600kHz separation as per the UK 2M Repeater Band Plan
  • Please see my note at the end of the list RE Desensitization
  • No, I don't yet have any filtering designed or implemented on the audio side, yet. At the moment, I've only got the audio from the MP3 directly going to the MIC IN on the transmitter via a 0.1uF series capacitor for DC Blocking.

RE: De-sensitising:-

That's a very good question and at the minute, the only issue I've identified (until I do more testing at least) is that the Receive Antenna will ideally need to be at least a wave length apart from the transmit antenna. I know that my local repeater uses large cavity filters - among other things - but I have to confess, I've not got that far ahead yet.

Since this is a portable project, I wanted small portable antennas that I could mount remotely from the repeater unit and keep them a wave length apart so I opted to go with two of these since they're small and portable. But an end user can use whatever they wish really.

I must confess, I started out just wanting to build an Amateur Radio Repeater but as I progressed, I started to see improvements and other potential applications where a small portable repeater could come in handy such as disaster relief in areas where there are no mobile networks and simplex radio signals are patchy / unreliable and so my design has grown with these influences to become something quite versatile and feature packed.

Would you believe that my first design concept prototype cost nearly £100 to produce initially? And while it cost that much, it didn't do anything more than the version I am working on today? However by comparison to where I am today it was large, clumsy, expensive and power hungry.
The early version used a complete walkie talkie handset and the morse was generated by the arduino and voice announcements were in WAV format on an SD Card which the arduino had to handle and play. This ate tremendous amount of prog mem (my code took up around 80 - 90% and the compiler would warn me of instabilities etc in the Atmega) and processing power but it proved the idea.
So from that point to this, my main focus has been to optimise what I had so far. To shrink it down in size and cost and make it more efficient. So where before in standby it would draw something silly like 30mA at standby, it now draws barely 2mA. I now use cheaper tiny transmitter modules instead of large expensive hand held radios and I've added power control via MOSFETS so only the Atmega is powered up all the while in sleep mode.
I think the cost of a prototype at this stage today is around £30 to produce. Ideally I'd like the entire completed project to cost no more than £50 to produce not including the cost of the battery or antennas.
And prior to all that horrible interference I was getting with the previous transmitter modules I was using, I had just designed in a small SIM800L Module for remote control and to issue telemetry such as battery charge level via SMS back to the user in areas where there is mobile network coverage. These modules only cost £8 and only one is needed per unit. I am also toying with the idea of including some sort of on board wifi as well - or maybe allowing for it to be included as a separate expansion.

Hope this helps :slight_smile:

Regards,

Hi;

Half Duplex for now - only one person may transmit at a time while others listen

That is Full Duplex, you receive and transmit at the same time.

Do you require a permit for repeaters in UK?
We do in Australia, even on the Amateur Bands as a callsign has to be registered.

Check with RAYNET, I'm sure they have portable repeaters too.

Tom... :slight_smile:

TomGeorge:
Hi;
That is Full Duplex, you receive and transmit at the same time.

Do you require a permit for repeaters in UK?
We do in Australia, even on the Amateur Bands as a callsign has to be registered.

Check with RAYNET, I'm sure they have portable repeaters too.

Tom... :slight_smile:

Hi, Yes in the UK to run an Amateur Radio Repeater you must be a full license holder to apply for a NoV from OfCom (The Office of Communications - which are the UK Spectrum Authority)

In the UK we have to go through a progressive license. We start at foundation level which permits you to transmit on 10W with the exception of some bands and frequencies and the course gives you an "appreciation" of the basics such as some theory, some practical work etc.

The next level is called intermediate which is slightly more advanced and that builds on the knowledge gained through the foundation course. This level introduces some of the formulas used, more advanced design and build knowledge and will entitle you to use up to 50W of power and a few more privileges.

And finally we have the full license which is really in depth and of course builds on the previous two levels. Once passed, you can use up to 400W (the max legal UK limit) and run remote unmanned stations and a whole lot more besides.

Hi,
What level of licence do you have.

Tom... :slight_smile: