I need a PRO's opinion on this circuit please

[lilimike]

Hi,

I would like to get a Pro's opinion on this circuit. It is the receiver part of an ultrasonic distance measurement device. The Tx is performed straight from 2 PIC pins (inverted) with a burst of 40KHz duty cycle of 50% for a delay of 150mS.

I am using 400ST160 and 400SR160 transducers. I have tried to improve the gain but it doesn't seam to make a difference. Maybe I need to work on the Tx part by using a MAX232 but I would like to make this my last option. If I could get some good advice on the receiver by adding or removing parts or changing values that would be great help.



Thanks

Mike

 

[mneary]

I wouldn't normally put a DC bias a piezo transducer, and 3.3k seems like a pretty low impedance. What does the data sheet suggest?

 

[lilimike]

I am not sure how to interpret this part on the datasheet.

Here is the link to it

 

[crutschow]

The transducer has maximum sensitivity for a load of 100kΩ or higher.

 

[lilimike]

Sorry Carl,

Now I am not sure how to interpret what you are saying. Do you mean that I should replace R4 with 100K or higher?

 

[crutschow]

Yes, R4 should be 100kΩ. Higher would not give any advantage.

Also, as mneary noted, I don't see any reason for bias on the transducer. Connect the transducer across R4.

The input bias current offset of U3-B could be a problem. The 200nA offset through the 100kΩ resistor gives a maximum output offset of about one volt at the amp output, which affects the sensitivity of the LM393 comparator. To eliminate this effect you could capacitively couple the output of U3-B.

Why is the trigger point of U5-A comparator set at 1.6V?

 

[lilimike]

100KΩ for R4 with transducer connected across it is working well.

Should I couple the output of U3-B with .01µ or .1µ ?

The trigger point of U5-A at 1.6V was not based on calculations, I just measured the best trigger with a scope using a trim pot and when it was at best I replaced the pot with 2 resistors.

 

[crutschow]

If you adjusted the trigger point experimentally, then you likely took into account the affect of any offset, so you may not need a capacitor.

What is the DC output from U3-B?

 

[lilimike]

U3B output is 8.2VPP
In this graph bellow CH1 is triggered PWM from Tx and CH2 U3B output
The echo at 10mS is the sealing of my lab at apr. 5.5 feet and the first echo I figured is the vibration caused by the Tx when it transmits the burst

 

[ericgibbs]

U3B output is 8.2VPP
In this graph bellow CH1 is triggered PWM from Tx and CH2 U3B output
The echo at 10mS is the sealing of my lab at apr. 5.5 feet and the first echo I figured is the vibration caused by the Tx when it transmits the burst
View attachment 42611

hi,
Whats the output from U5A the comp look like.?
Whats the TX time period for the 40kHz.?

 

[lilimike]

Hi Eric

On the graph shown I transmit 40KHz for 200 µs, I have tried playing with this value and found the best performance was between 150 to 200

Here is a picture of the same setup as above with CH2 at U5A

 

[ericgibbs]

Hi Eric

On the graph shown I transmit 40KHz for 200 µs, I have tried playing with this value and found the best performance was between 150 to 200

Here is a picture of the same setup as above with CH2 at U5A
View attachment 42613

hi,
That TX pulse width is about right for 40kHz, its a good looking echo after the comparator.

With regard to the ringing of the transducer after the TX period, this is quite normal.
Usually you would have a suppression gate on the receiver for a period after the TX pulse.

Also depending upon the ranges and accuracy you may require TGC, [timed gain control] so the gain of the receiver increases with time after the transmit pulse.

 

[lilimike]

Thank you Eric,

I control the ringing by software so I do not listen to the echo until a short delay after I transmit. (still playing with values)
As for TGC I think this extra circuit will not be needed, the further the distance, the less accuracy I need and as it is now within 6 feet I get about half an inch of accuracy which is more than acceptable.
I am now at the stage where I am satisfied with the tests in my lab and that is between my bench and the ceiling. I now have to change my testing environment so I will move my lab to the garage and start testing distances horizontally and of course I still need to bring in some different conditions such as cold and heat and loud noises.

 

[Roff]

The input bias current offset of U3-B could be a problem. The 200nA offset through the 100kΩ resistor gives a maximum output offset of about one volt at the amp output, which affects the sensitivity of the LM393 comparator.

200nA*100k is only 20mV, and this is not amplified by the gain, which is unity at DC, due to C11.

 

[lilimike]

Hi Ron,

Are you suggesting that I change something? or confirming the values are ok?

Thanks

Mike

 

[Roff]

Hi Ron,

Are you suggesting that I change something? or confirming the values are ok?

Thanks

Mike

I am only suggesting that, with the values shown, offset should not be a problem on U3B. I agree that R4 needs to be 100k.

 

[crutschow]

200nA*100k is only 20mV, and this is not amplified by the gain, which is unity at DC, due to C11.

Your are correct of course. I forgot that the capacitor reduces the DC gain to unity for the offset current.

 

[lilimike]

One more question...(among thousands of others)
To power my circuit which has now become this:

I was planing to use 2 regulators to supply 5V and 9V and use an 12 V AC/DC plug-in type as main power like this:

Since the 9V is only for LM833 and LM393 which both support 2V to 16V, would it cause trouble if I got rid of U4, C4 and C5 and use as main power an unregulated 9V 100mA+?
I would save a few pennies on components and a few cm² on PCB space!

Thanks

Mike

 

[mneary]

The 9V also supplies bias to the microphone. Filtering (R5, C7) may need to be improved.

It also provides a reference to the comparator (R13, R14).

 

[lilimike]

With the reference to the comparator between R13,R14 can I assume that the voltage there is always approx 18% of the supply voltage so it should not matter as long as it is always proportional as well as R5, R6 ?

As for R4, C7 what is the calculation to find the best values?

I have tested with using the 7809 reg and I get:
1.6V at pin 3 of U5
4.4V at pin 3 of U3
128 mV across Rx

I switched to a 9V plug in adapter and my values are now:
16V from the adapter's main at no load
10.4V with load (my circuit)
2.8V at pin 3 of U5
7.8V at pin 3 of U3
80 mV across Rx

I seam to be getting the same range from the Rx using both ways.

Another concern, if I want to mass produce my circuit, can I trust that each adapter perform the same or close enough?


Mike