Hi guys,

I have a LC tank that I can tune to some low freqs so I thought about building a receiver for a radio sation that transmits at 77.5Khz, it's this one:http://en.wikipedia.org/wiki/DCF77

In my inmense naivity I thought, beh the tank circuit is already resonating with whatever it receives from Germany so all I have to do is to amplify that. Wohoo!

Op amps are nice so I chose a non-inv configuration that gives me a high input impedance so the LC tank will resonate freely.

BUT MY circuit all it does is to oscillate at the freq of the LC tank and always with the same amplitude...

I'm sure I am doing something terribly wrong, could you guys help me?

Thanks!

 

Hello,

The following page might interest you:
http://www.marvellconsultants.co.uk/DCF/

There you will find some schematics for receiving the DCF77 signal.

Bertus

 

Hi,

Yeah, I found some circuits, but not knowing why my circuit doesn't behave as i'd expect is killing me It should be enough right? So why is it oscillating?

 

Hello,

There can be problems with decoupling.
Have a look at the following thread:
Decoupling or Bypass Capacitors, Why?

Also put a decoupling capacitor on the voltage divider, from the middle to ground.

Bertus

 

I added a decoupling cap between vgnd and gnd and it is still oscillating... Where else should I put decoupling caps?

Also:

- Is my design good? Is it fair to expect it to pick radio waves and get them ampilfied?

- Could the the output of the opamp radiate and affect the LC tank? How can I avoid this coupling?

I cant wait to see my radio working
Cheers!

 

Hello,

The LM324 has a to low gain for your circuit.
You have set a gain of 101, for wich a GBW of about 8 MHz would be needed.
The LM324 has a GBW of 1 Mhz.

You could try to lower the gain by reducing R4 to 10 KOhm and see what happens.

Bertus

 

Hi Bertus

I reduced that gain and the oscillation stopped Wohoo (I cant find in the data sheet where doest it to talk about the GBW BTW!) I had to use a second op amp from the lm324 to amplify even more then.

I used an AD9850 to make a fequence sweep and got a nice bode diagram (blue vertical lines). The AD9850 is not connected to the tank circuit I just have a 10cm wire plugged to its output that just radiates around.

I was expecting the LC tank would be better at selecting the frequency I am interested in! Am I being too exigent?

It looks like I need better filtering and then I need to amplify the signal even more.

Does it sound right?

 

Hello,

Here are two datasheets of the 324.
The TI sheet says 1.2 Mhz.
The national sheet says 1 Mhz.

Bertus

 

I added a decoupling cap between vgnd and gnd and it is still oscillating... Where else should I put decoupling caps?

Also:

- Is my design good? Is it fair to expect it to pick radio waves and get them ampilfied?

- Could the the output of the opamp radiate and affect the LC tank? How can I avoid this coupling?

I cant wait to see my radio working
Cheers!

Frankly,it doesn't look a lot like a radio at all in its present form.

Although Bertus' link does use Op Amps like yours it is a lot more complex.
It uses an Active filter instead of your LC circuit.

It also uses an Active Antenna circuit to increase sensitivity,as well as match the input impedance of the main Rx to that of the very short antenna.

It is also not representative of how Radios are normally made.

Why not look up some information specifically about radios?

Google for VK3YE,who has some really simple radios on his website,including a simple Software Defined Radio.

With your existing circuit,try shifting the LC tuned circuit to the input of the "radio".

It just might work,but will be horribly insensitive.

.

 

Hello,

A simple active antenna most times uses a fet as input transistor.
Here is an example of a schematic:

The image is from this page:
http://www.qsl.net/dl4yhf/dcf77_osc/index.html

Bertus

 

Hello,

A simple active antenna most times uses a fet as input transistor.
Here is an example of a schematic:

The image is from this page:
http://www.qsl.net/dl4yhf/dcf77_osc/index.html

Bertus

Yeah,& as an old RF "greybeard",that is what I would have expected.

The one on http://www.marvellconsultants.co.uk/DCF/
uses Op Amps,though----maybe special ones,but Op Amps nevertheless.

 

- I understand my approach is a bad one and I saw the one from qsl.net, but I'd really like to know why my design is a bad one. More than trying to build a receiver I am trying to understand why things are in a certain way.

- A FET is used because it's high impedance input, this allows the LC to resonate freely... But so my op-amp has high impedance! It should be good right? Why is it bad?

- I also reduced gain as Bertus suggested and I added a second stage inv-opamp (thanks Bertus!). The system stopped oscillating, I made a bode diagram by having an AD9850 make a frequency sweep, note the spike at the 77.5Khz (more or less) I was expecting to see a higher selectivity, why does it have such a low sensitivity?

- Why are op amps bad for radios? Is it because of the BW? Are they also bad for 77.5Khz?

Thanks!

 

Well,I can't pretend to be an Op Amp Guru---the last time I did any serious analysis of them was nearly thirty years ago,sandwiched in between Differentiation of Trignometric Expressions & Microprocessors.

If I remember correctly,we didn't go very deeply,unlike FETs & BJTs where “we went right down to the bone”!
Since that time,most of the Op Amps I've encountered have been well behaved beasties,but there have been a few occasions where I've run into problems with instability.

As Op Amps rely upon quite heavy levels of negative feedback to operate in the linear mode,reactive loads,or reactive components in the feedback circuit can add enough phase shift at some frequencies to turn the feedback positive,with resultant oscillation.

More on that later,let's have a look at your circuit:-

I think you said you have an antenna about 10m long----at 77.5kHz that is 0.0026λ.

Very short antennas usually have a radiation resistance of only a few Ohms,most of the Impedance being capacitive.
The normal way to try to match this impedance is to have a tuned LC at the receiver input with either a small coupling winding wound over the inductor,or a series of taps on the inductor.

The input impedance of your circuit is effectively R1 in parallel with R?,giving you 50kΩ.
( To AC,V1 supply looks like a short circuit ,so both 100kΩ resistors are in parallel)

Not a spectacular match!

Next we have U2,which normally has no voltage gain,but does contribute noise,if nothing else goes wrong.

L1 C1 appear as part of the load for U2,so can possibly cause the feedback reversal referred to earlier.
(This was my first thought,but from your later comments,it seems more probable that U1 is the culprit !)

L1C1 are not directly in U1's feedback path,but are only isolated by R5,so it looks possible that's where you are getting positive feedback,but then,as I said,I'm not an Op Amp Guru!

Going on,out of three active devices,if everything else worked OK,you've only got a voltage gain of 100—not marvellous,when you are probably only starting out with a few tens of μVolts!

Why general purpose Op amps aren't used at RF:-
VLF radios are quite rare,& most of the design progress in this field is driven by HF & above.
Most early Op Amps had severe bandwidth limitations,& were not suitable for use in 455kHz IF stages of MF Superhets,let alone in the front ends of such receivers.

Even those few with reasonable gain/bandwidth factors required dual supplies,& offered no reduction in parts count compared to discrete circuitry,as either extra discrete components were necessary to provide AGC,or several of the Op Amps would have to be dedicated to this.

Compare this to the ease of providing AGC to a dual gate FET stage!

Many early Op Amps would oscillate “at the drop of a hat”,& required external stabilisation circuits.
Even the ones with inbuilt stabilisation were still prone to instability if operated into reactive loads.

Eventually,Op Amps improved to the point where they might have been useful,but by that time many discrete radio circuits were replaced by dedicated ICs which included several circuit functions in one chip,not much larger than a single Op Amp.

 

Besides, building anything with high gain on a plug-board is asking for trouble (can you say oscillations?)/

 

Besides, building anything with high gain on a plug-board is asking for trouble (can you say oscillations?)/

Very valid point---I hate those things!

 

Hi!

Just to clarify,

V1 is just a DC source of 5 volts, and U2 is just there to provide a virtual ground of 2.5v.

My antenna is just a ferrite coil (no 10M antenna) that should be able to resonate freely because of the high impedance input of U1. I am now using 2 op amps with reduced gains instead of just one with a huge gain... that got rid of oscillations though.

I hope this helps

 

Hi!

Just to clarify,

V1 is just a DC source of 5 volts, and U2 is just there to provide a virtual ground of 2.5v.

My antenna is just a ferrite coil (no 10M antenna) that should be able to resonate freely because of the high impedance input of U1. I am now using 2 op amps with reduced gains instead of just one with a huge gain... that got rid of oscillations though.

I hope this helps

Yeah,reading back,it was a 10cm antenna on your signal source----sorry about that!

It makes a bit more sense,now--pity you didn't make it a bit clearer in your first posting.
Your chosen schematic program tends to be a bit hard to read---normally U2 & associated circuitry would not be drawn directly in line with the signal handling circuitry.

You should really connect some bypass capacitors between your virtual ground & the two power rails,so it looks more like a ground for RF.

This,however,raises the spectre of a reactive load for U2,so it may be an idea to put a low value resistor in series with its output pin.

My comments re the use of Op Amps in radio receivers still apply,& I think you may be battling to get satisfactory results.

 

Strangely enough,when I Googled for "Op Amps at RF",I found several enthusiastic articles about the use of Op Amps in RF applications.

They all used higher performance devices than the LM324,& concentrated on wideband circuits,not conventional LC tuned amplifiers.
Where they did use selectivity,it was obtained using SAW filters,or ceramic filters.

As I say,they were all very enthusiastic,but they were comparing "Op Amps with discrete circuits.
There may be some advantages,but in my opinion they are minimal,& in any case,as I said earlier,most modern design uses "purpose designed" multifunction ICs.

Such devices lend themselves well to the dominant Receiver (& in many cases,Transmitter) Superheterodyne technique.

More recent Direct Digitisation receivers are more likely to use FPGAs.

It seems that,in RF use, apart from a few "niches",OpAmps have left their run a bit late!

 

Sorry,double posted--couldn't get it to delete