40KHz AM Demodulator

[arhi]

Hi,

I have a need to demodulate the 40KHz ultrasound signal (that I generate on the other side using uC). In order to make things more clear, lemme spend some words explaining the circuit.
- block 1 is uC and other digital elco that among other things generate 40KHz chirps via 40KHz ultrasound transducer (UST40T)
- block 2 is receiving sound from 40KHz ultrasound transducer (UST40R) demodulating it and sending digital and analog signals to uC in block1

The idea is, b1 send chirp, chirp bounces back from obstacle, b2 catch the chirp and uC measure the time ... now, in theory this all works perfectly, in practice, I'm terrible at analog elco so this thread is actually about 2 things

1. Is there any pandam to IR Receiver in ultrasound world (one/two parts that will demodulate 40KHz signal and return TTL output)

2. if answer to 1 is NO, and I presume it is, can someone look at the attached schematic (drawn by me, calculated / developed by someone else) and comment on - do I need to recalculate some values (capacitors mainly), do I need to use another type of OP (LF353N is dual wide bandwidth opamp with jfet input, I have no idea how "new" or good it is, never used it). I understand how the schematic works that is not the problem, just not sure about quality of the opamp and calculation of the R/C values.

I was hopefully clear so .. thanks in advance

 

[audioguru]

Hi Arhi,
The output of your circuit is a signal that swings from -10.5V to +10.6V. It will need to be attenuated and level-shifted to be from 0V to +5V for the micro-controller.

The first opamp is inverting with an input impedance of only 4.7k. I hope that your receiving transducer can drive it in parallel with the 10k input resistor.

The high values of your coupling capacitors pass audio frequencies to below 400Hz. Hopefully the transducer will not pick up voices.

The older FET-input opamp might have the problem of Phase-Inversion where the output suddenly inverts when the input level is high.

 

[arhi]

thanks for the hint .. I recon that I can make the output swing from -2.5 to 2.5 and then shift it up if I power the opamp with lower V, and then shift that to 0-5V, but the most important part of your answer is 400Hz part ..

As I mentioned, I suck terribly at analog elco .. I know the basics, and that's where it stops.

The transducer have pretty nice curve for sensitivity, I will attach the datasheet. And as for the 10K input, I will play with it when the transducer finally arrive (I hope tomorrow).

Thing is, I do not have to use this opamp, I do not even need to use this schematic but this was the best I could find searching the net / my old notes .. And I do not understand the 400Hz part, do you mean that everything above 400Hz will go trough this circuit, if yes, then the basic point of the circuit is gone

The idea is that I send a 40KHz signal that is 0.0025 sec long and then wait for it on the other side, measure time when I start receiving it and measure time when I do not receive it any more (start and end of the chirp). So I'm not "packing" some real modulation in the 40KHz it will just be "on/off" ... The whole point of this circuit is to eliminate all other noise so to receive only 40KHz signal ... something like this:

signal sent from block1:
     _   _   _   _   _   _   _   _   _   _
____| |_| |_| |_| |_| |_| |_| |_| |_| |_| |___________________________________


signal received from transducer (input of block2):
                             _   _   _   _   _   _   _   _   _   _
____________________________| |_| |_| |_| |_| |_| |_| |_| |_| |_| |___________


output from block2:
                             _____________________________________
____________________________|                                     |___________

 

[arhi]

audioguru, every suggestion is welcome .. here is the list of the opamps I can get here so you might recommend better one then the one on the example

739, 741, 747, 748,
CA3054
ICL7650SCPD,
L272, L272M,
LM10CN, LM13700N, LM201AN, LM224N, LM258N, LM2902N, LM2904N, LM301AN, LM308N, LM318N08, LM324D, LM324N, LM348N, LM349N, LM358D, LM358N, LM386M, LM3900N, LM392N, LM4250CN, LM725CN, LM776CN, LM833N, LM837N,
LP324N,
LT1014CN,
M5216L, M5238P
MC1458D, MC1458N, MC1558N, MC33078N, MC4558CN,
NE5532AN, NE5532N, NE5534AN, NE5534N,
NJM4558L,
RC4136N, RC4559N,
TAA981,
TDA1029

that's about it .. that is full collection of opamps I can get in local elco stores , yes I know it is patetic but .. with most suppliers unwilling to ship to this country, I have to make due with what I can find locally.

Anyhow, thank you for looking at the schematic, if you believe you see better part for the job, please, point me to it, and if you believe this schematic will "get the job done" please tell me so, also if you think otherwise, a nudge in right direction will help ... I do suck at analog but I can read and I'm not lazy

 

[audioguru]

As I suspected, the low input impedance of the first opamp loads down the receiving transducer and reduces its output.

The formula for the -3dB response of a coupling capacitor is 1 divided by 2 x pi x RC.
Other ultrasonic circuits use a 1000pF (1nF) input capacitor to reduce the low frequencies.

Use the LM833 dual opamps for this circuit.

 

[arhi]

thanks. I knew some of the op-amps sold here can be used (is lm833 generally ok? or it is just best for the job from the patetic list I wrote here?)

I will try to recalculate the values, who knows I might be even able to simulate this in multisim or isis

thanks again

 

[audioguru]

The LM833 is also wideband and has low noise. I don't think it has the phase inversion problem. Use it.

Here is a Japanese circuit similar to yours: Ultrasonic Range Meter