long wave 198khz front end tuner using active band pass filter op amp

Thread Starter

davidjohnhills

Joined Oct 8, 2017
13
Good Morning

Could someone tell me if it is feasible to build a front end tuner for long wave (say 198khz radio 4) using a op amp as active filter band pass.

It is going to connect to a long wire antenna.(Input) and crystal ear piece (output)

I have a SparkFun Configurable OpAmp Board - TSH82

with has a 3 dB-bandwidth: 100 MHz
the filter needs to have bandwidth of 9khz

I would like to get some gain from it , perhaps x10 to x50

and configure it to work as a precision rectifier at the same time.(not so important, if I can get some voltage amplification)

link to data sheet
https://cdn.sparkfun.com/assets/d/d/3/a/0/tsh82.pdf

many thanks

David H
 

danadak

Joined Mar 10, 2018
4,057
A quick and dirty way of doing this would be to get a 455 Khz IF transformer
off ebay or Digikey, add C to the primary and secondary to lower its resonance.

That gives you a 4'th order transfer function which should approach the Q you
are seeking.

Follow that up with the OpAmp to get the G you want.

1576161089488.png


Representative part - https://www.minikits.com.au/doc/Xicon-42IF104-RC.pdf


Regards, Dana.
 

Thread Starter

davidjohnhills

Joined Oct 8, 2017
13
Thanks for your reply Dana

Could I use a couple of these in series? which I have a surplus of.


TRANSFORMER, FOR ULTRASONIC TRANSDUCER, K4000005 1774948
You are purchasing part number K4000005 Manufactured by PROWAVE
TRANSFORMER FOR ULTRASONIC TRANSDUCER
Inductance : 1.5mH
Product Range : -
SVHC : No SVHC (17-Dec-2015)
External Depth : 7.9mm
External Length / Height : 10mm
External Width : 8.3mm
Frequency Max : 125kHz
Lead Diameter : 0.5mm
Lead Length : 2.5mm
Lead Spacing : 4.5mm
Transformer Mounting : Through Hole
Transformer Type : Matching
Turns Ratio : 1 : 20
 

Thread Starter

davidjohnhills

Joined Oct 8, 2017
13
Band-Pass; 2nd order Butterworth; 9kHz passband; 198kHz center
Good evening
Would this work(at all) and with a TSH82 instead of the AD8033?
thanks all


bode-.png
filterwizard-[1].png
 

danadak

Joined Mar 10, 2018
4,057
Certainly appears to do so in the sim, using the OpAmp approach.

If the sim using 8033 produced the plot, then THS82 should work because THS82
has more GBW.

Bypassing with these fast OpAmps critical, both ceramic and tant would be in order.

Type of caps with good ESR -

1576173800855.png


Pay attention to component tolerances, the OpAmp filter approach caps getting small enough
where layout and stray C can affect. Since this is a high Q design all the more important. Note
the transformer with tuned slug, is potentially more accurate, but requires manual tuning.



Regards, Dana.
 
Last edited:

jpanhalt

Joined Jan 18, 2008
10,066
What you show is known as a Delyiannis-Friend circuit with Q enhancement or more simply, a single amplifier biquad (SAB). The Q enhancement comes from positive feedback (i.e., the 6.97k and 121R0 resistors). I would simply build it up on protoboard. I prefer the solderable type, but the non-solderable will probably work too. Modifying your SF board would be a pain.

You may not need nor want Q enhancement. It can give a narrow response but can also produce a nice oscillator at that frequency.

If you want to go deeper into the design than the design tools, I recommend "Design of Analog Filters" by Schaumann and Van Valkenburg, Oxford University Press ISBN 0-19-511877-4, starting about page 170. I believe it is the second edition, not the later one with 3 authors, and is available online. You might also want to download SLOA096 from TI.
 

danadak

Joined Mar 10, 2018
4,057
Here is a sim of the filter, note 8034 is a dual of the 8033.

1576234925634.png


The accuracy of the resonant peak basically dependent on R & C tolerances, T & V. So if thats
critical you will need adjustment, the IF transformer being a more stable approach.


Regards, Dana.
 

Thread Starter

davidjohnhills

Joined Oct 8, 2017
13
Thanks for your help
If I don't use the THS82 what is the minimum GainBand width Product and slew rate to make this stable (gain say V/V 20 to 30)

Is there a simple way (rule of thumb) of determining the min safe required GBP.

for example the THS82 has a 3 dB-bandwidth: 100 MHz could I use this with a gain of 10 at 10Mhz?
 

danadak

Joined Mar 10, 2018
4,057
Thanks for your help
If I don't use the THS82 what is the minimum GainBand width Product and slew rate to make this stable (gain say V/V 20 to 30)

Is there a simple way (rule of thumb) of determining the min safe required GBP.

for example the THS82 has a 3 dB-bandwidth: 100 MHz could I use this with a gain of 10 at 10Mhz?
If the OpAmp has ~ a single pole response yes.

https://en.wikipedia.org/wiki/Gain–bandwidth_product


Regards, Dana.
 

danadak

Joined Mar 10, 2018
4,057
I have a few ne5534ap they have
Unity-Gain Bandwidth 10 MHz Typ
High Slew Rate 13 V/μs Typ

Would they be Ok (how do I work it out)
thanks
David

http://www.ti.com/lit/ds/slos070d/slos070d.pdf
I would try it.

To work this out either a sim (simplest method) or an actual LaPlace analysis
of circuit would be in order. Approximations would have to be used as that
would be a 4 pole analysis typically, 2 for OpAmp, 2 for RC networks.

The LaPlace analysis a bit ugly algebraically but one pole is far removed from the
other 3 so that turns into a 3'rd order analysis = a little easier.

But the 5534 OpAmp GBW limits G to ~ 10.


Regards, Dana.
 
Top