TL072 questions

[audioguru]

OK, I have looked up the MC3407x and it looks like they're discontinued.

No.
Digikey has 1016 MC34071APG in stock for $1.20US each. The A is the premium one and the P is a DIP package.
They also have 1184 MC34072APG in stock for $1.07US each.
They don't have any MC34074 in stock and don't have any MC3307x in stock.
In Ontario the stuff I order from Digikey before 8:00PM is delivered the next morning.

A suggested replacement is the TLE214x, which seems to fit the bill, except it doesn't have FET inputs so I might need to go to the non-inverting amp after all (not sure about this yet).

Your inverting amplifiers don't need a FET input. Their input impedance is only 10k ohms. The TL08x and TLE214x have FET inputs and the MC3407x has transistor inputs but the type of input doesn't matter.

1) I think that the voltage on the base of Q1 does not suffer limiting from D1 because of the base-emitter diode action of Q1. Is this true?

The diode is connected in the opposite direction to the base-emitter diode. The diode allows the 22nF input coupling capacitor to charge and discharge equally. Without the diode then the capacitor would charge then not couple signals anymore.

2) I think I'm OK pulling the bias on the positive input of U1 so hard to Vcc because I *want* to slam the output from 5V to 0V when receiving. Is this reasonable?

No.
The inputs of these opamps don't work when they are within 2V from the positive supply. Most opamps that have FET inputs suddenly invert when an input is within 2V from a supply. The outputs have a swing of about 1V to 4V.

 

[Roff]

Not that it matters, but the TLE214X has PNP input transistors.

 

[audioguru]

Not that it matters, but the TLE214X has PNP input transistors.

You are correct, Ron and I was wrong. I thought they had Fet inputs. I was thinking about the OPA134, OPA2134 and OPA4134 opamps with FET inputs.
The MC3407x also have PNP inputs so the inputs can still work at the negative supply voltage which is ground in this circuit.

 

[Torben]

You have biased the input at 4.5V, when the input common mode range extends to only 3V. You should probably set the input bias voltage at about 2V.
You have a highpass filter (220pF, 10k) on the input at 72kHz. Your input capacitor should be at least 1.6nF, which would set the corner at 10kHz.
EDIT: I see why you picked the values you did. In the simulation, it works with them, and it doesn't with the values I recommended.
I'm not sure I would trust the spice model. You are clearly exceeding the common mode range by a large amount. You can probably get the circuit to work, but you should probably be prepared to juggle component values when you build hardware.

Gotcha--I wondered what the effect of the input cap and resistor might have in that regard but I didn't recognize it as a filter.

Are you suggesting 2V instead of closer to the limit so that it's conservative and less likely to bump the edge of the voltage range and risk inverting the phase?

As for the simulation, that would have been my third question, had I remembered to ask: I suspect simulations are better if one understands why they might be wrong and has the ability to double-check. I'm trying not to use it as a crutch for learning but it's very seductive. I've already witnessed a couple of circuits which performed differently on the bench than in the computer.

Thanks for the help!

Torben

 

[Torben]

No.
Digikey has 1016 MC34071APG in stock for $1.20US each. The A is the premium one and the P is a DIP package.
They also have 1184 MC34072APG in stock for $1.07US each.
They don't have any MC34074 in stock and don't have any MC3307x in stock.
In Ontario the stuff I order from Digikey before 8:00PM is delivered the next morning.

Great! Of course now I can't find the source that made me think they were discontinued.

Now I'm trying to get enough projects together for the next little while to meet the $32.50 CDN minimum order where the service charge is waived.

Your inverting amplifiers don't need a FET input. Their input impedance is only 10k ohms. The TL08x and TLE214x have FET inputs and the MC3407x has transistor inputs but the type of input doesn't matter.

You're right, of course--I wasn't thinking about what you said earlier. (Noted: the TLE214x have PNP inputs, and this is irrelevant anyway, as you and Roff have already agreed).

The diode is connected in the opposite direction to the base-emitter diode. The diode allows the 22nF input coupling capacitor to charge and discharge equally. Without the diode then the capacitor would charge then not couple signals anymore.

OK, that makes sense.

No.
The inputs of these opamps don't work when they are within 2V from the positive supply. Most opamps that have FET inputs suddenly invert when an input is within 2V from a supply. The outputs have a swing of about 1V to 4V.

OK, Roff noted this as well, and now that I look at the data sheets (and AoE, and Op Amps for Everyone, etc) more closely, it makes sense.


Thanks,

Torben

 

[Torben]

Hi all,

OK, where I'm at now I think I'm OK with:

1) The TL072 can work at 5V, but not very well, and it's below specs (although some graphs on the datasheet show that it works but starts degrading a little below 5V for what I need). So 2 stages of TL072 work for this at 5V, but not just one--it just can't operate well enough at 5V.
2) A single stage TLE2141 can do what I need clear down to 3V. However, the output square wave is not 50% duty cycle--that is, after the BC550, it doesn't stay at ground as long as it stays at 5V. Since the 8051 is detecting the ground condition, I want longer grounded times. (The dual TL072 has about a 50% duty cycle output).
3) Putting another BC550 configured as a common-emitter amplifier in front of the TLE2141 as a preamp cures the problem in 2) above. But if I'm going to do this I might as well use 2 stages of the TL072 or TLE2141 anyway.
4) The TL072 also does the job in one stage with the preamp attached.
5) This is all simulated and on paper with a calculator, so real-world performance is still to be tested.

So I'm going to order some stuff (MC3407x, TLE214x, some 1% resistors, etc) from Digikey as soon as I figure out what else I want to bulk up my order, and test the different configurations to see what I like best. Also I'm going to revisit my requirements--the 8051 software as it is doesn't need to know what the frequency is, just that it's been received. But I'll think about that more when I revisit the software itself. Right now it's pretty stupid and is just a counter waiting for an input to go low. I'm going to look into rejecting early reflections etc, which might need me to have the frequency available to the program. At that point this project will probably wind up on the microcontroller forum.

Anyway, that's just a thread recap of where I am now. Thanks for the help to get me this far!


Torben

 

[Roff]

Torben, have you considered just using a comparator? If you don't use the other comparator section, you need to tie one input only to GND.
Below is the model for LM393, if you don't have it. Save it as LM393.sub in the SUB folder under LIB.

* LM393 VOLTAGE COMPARATOR "MACROMODEL" SUBCIRCUIT
* CREATED USING PARTS VERSION 4.03 ON 03/07/90 AT 14:17
* REV (N/A)
* CONNECTIONS:   NON-INVERTING INPUT
*                | INVERTING INPUT
*                | | POSITIVE POWER SUPPLY
*                | | | NEGATIVE POWER SUPPLY
*                | | | | OPEN COLLECTOR OUTPUT
*                | | | | |
.SUBCKT LM393    1 2 3 4 5
*
  F1    9  3 V1 1
  IEE   3  7 DC 100.0E-6
  VI1  21  1 DC .75
  VI2  22  2 DC .75
  Q1    9 21  7 QIN
  Q2    8 22  7 QIN
  Q3    9  8  4 QMO
  Q4    8  8  4 QMI
.MODEL QIN PNP(IS=800.0E-18 BF=2.000E3)
.MODEL QMI NPN(IS=800.0E-18 BF=1002)
.MODEL QMO NPN(IS=800.0E-18 BF=1000 CJC=1E-15 TR=807.4E-9)
  E1   10  4  9  4  1
  V1   10 11 DC 0
  Q5    5 11  4 QOC
.MODEL QOC NPN(IS=800.0E-18 BF=20.29E3 CJC=1E-15 TF=942.6E-12 TR=543.8E-9)
  DP    4  3 DX
  RP 3  4 46.3E3
.MODEL DX  D(IS=800.0E-18)
*
.ENDS

 

[Torben]

Torben, have you considered just using a comparator? If you don't use the other comparator section, you need to tie one input only to GND.

Well dang. That looks like it might do the trick quite nicely. Thanks!


Torben