2
\$\begingroup\$

I'm trying to do on/off keying (OOK) to transmit messages with ultrasound.

This is the transducer.

This is the schematic of the signal generator and amplifier:

Schematic of the signal gen and amp

It seems that the class A amplifier doesn't work.

  • Have I made an error with the operating point?
  • Should I instead use opamps?

If you know how to fix this approach or if you have any better ideas please let me know.

\$\endgroup\$
6
  • \$\begingroup\$ I think you'll need to actively drive it "push-pull" and your example CE amplifier won't do that. I think that's one problem among several. \$\endgroup\$
    – jonk
    Aug 28, 2020 at 8:15
  • 3
    \$\begingroup\$ "Doesn't work" how? No output? Distorted output? Output level too low? It gets up off the work bench and kicks your cat? What did you expect, and what did you get? How did you check the output? How do you know that the driving signal is OK? \$\endgroup\$
    – JRE
    Aug 28, 2020 at 8:16
  • \$\begingroup\$ Haben Sie auch ein englisches Datenblatt? \$\endgroup\$
    – jonk
    Aug 28, 2020 at 8:17
  • 4
    \$\begingroup\$ Mmmpf. You won't get 40V output from a class A amplifier with a 5V supply. \$\endgroup\$
    – JRE
    Aug 28, 2020 at 8:18
  • 2
    \$\begingroup\$ You can edit those into your question. Comments get deleted sometimes. \$\endgroup\$
    – JRE
    Aug 28, 2020 at 8:45

3 Answers 3

5
\$\begingroup\$

If you are going to use a class A amplifier on such a low voltage supply then might I suggest you configure the collector to drive a step up transformer. You can arbitrarily deliver what ever voltage you need to the secondary (within reason) but you might need some heat sinking on the output transistor.

The next problem is delivering a sine wave to the load. At the moment, the 555 produces a square wave and that’s roughly what you’d get in your current circuit (with or without an output transformer) so, you might be advised to use an op-amp low pass filter between 555 and the amplifier. You could use a pi network filter of course if you knew how to design it.

It’s unclear about how you would intend to perform OOK in the circuit but I would think that some logic input to the 555 and another transistor could be persuaded to give this functionality.

\$\endgroup\$
19
  • \$\begingroup\$ Yes, the OOK part is not included in the shematic because I wanted to focus this question on the carrier frequency gen and amp. \$\endgroup\$
    – le_lemon
    Aug 28, 2020 at 8:47
  • \$\begingroup\$ Yes, later I will attach RC or Pi Filter inbetween the 555 and the amp. \$\endgroup\$
    – le_lemon
    Aug 28, 2020 at 8:48
  • 1
    \$\begingroup\$ What do you think about just an emitter circuit and a 1:10 transformer (like this mouser.de/datasheet/2/597/ga3459-463457.pdf) after it ? \$\endgroup\$
    – le_lemon
    Aug 28, 2020 at 9:41
  • 2
    \$\begingroup\$ @Andy aka: That got lost in translation. German Emitterschaltung means the emitter is tied, so it's the same as English common emitter circuit. \$\endgroup\$
    – Janka
    Aug 28, 2020 at 10:06
  • 1
    \$\begingroup\$ Yes ! The project just changed so much since and isn't finished yet so I didn't think about closing this question :D sorry \$\endgroup\$
    – le_lemon
    Sep 12, 2020 at 16:36
4
\$\begingroup\$

Your DC blocking capacitor forms a high pass filter in combination with the input impedance of the amplifier.

The input impedance is less than 100 ohms, giving a cutoff of over 200kHz.

Your DC blocker is filtering out your ultrasound signal.


Other problems:

  1. A simple class A amplifier can't put out a voltage higher than its supply voltage. You have 5V in, so 40V out is not possible with this design.
  2. The amplifier presents a heavy load to the 555. The 555 may have trouble delivering enough current to the BD139 base.
  3. You've already mentioned that you know that your signal is a square wave and that you'll have to filter it to make a sine wave out of it. Keep the low input impedance of your amplifier in mind when you design the filter. You might do better to useva buffer between the filter output and the amplifier input.
\$\endgroup\$
2
  • \$\begingroup\$ How would you solve those problems: should I increase the dc block capacitance ? Would you exchange the class a amplifier for something else ? I could attach a emitter circuit inbetween the 555 and the amp to lower the amount of current the 555 needs to deliver. \$\endgroup\$
    – le_lemon
    Aug 28, 2020 at 9:16
  • \$\begingroup\$ The DC block capacitor needs to be larger. Say, 100nF or more. An emitter follower would be one way to improve things, but I think you're going to want to use a different type of amplifier as Andy aka says. That also changes the DC blocking and what kind of buffer you might need. A class A amplifier wastes over 75% of the consumed power - it gets hot. \$\endgroup\$
    – JRE
    Aug 28, 2020 at 9:26
3
\$\begingroup\$

Ultrasound transducers require only very small currents. Instead of using an audio transformer, you might want to try a CMOS H-Bridge running at 15V. The H configuration gives you 30Vss accross the transducer.

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
7
  • \$\begingroup\$ If I understand you correctly, I just use 3 common emitter circuits because those act as inverters; on one side of the transducerr 1 and on the other 2 so I get for example -5V on one side and +5V on the other so that in total i get a Vss of 10V across. Like this: imgur.com/SPiuEsZ \$\endgroup\$
    – le_lemon
    Aug 28, 2020 at 12:06
  • \$\begingroup\$ In falstad this leads to a switching between -5V and +5V but not to 10V tinyurl.com/y4ra9whd. What did I do wrong ? \$\endgroup\$
    – le_lemon
    Aug 28, 2020 at 12:21
  • \$\begingroup\$ @le_lemon: Switching between -5V and +5V is 10V peak to peak. The transducer spec sheet lists peak-to-peak voltage. \$\endgroup\$ Aug 28, 2020 at 16:30
  • \$\begingroup\$ @DietrichEpp Thanks! Im gonna build this on monday and tell you how it went :) \$\endgroup\$
    – le_lemon
    Aug 28, 2020 at 17:54
  • \$\begingroup\$ No, my idea was to use a CD4049 inverter chip and run it at 15–18V supply voltage. That's within its spec. (They burn out at roughly 20V, so don't try that.) \$\endgroup\$
    – Janka
    Aug 29, 2020 at 1:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.