TL071 vs. TL072

[Paul Marossy]

I have to ask this question. Is a TL071 interchangeable with a TL072 in circuits that only use 1/2 of a TL072? Is is not true that a TL072 is a dual version of a TL071? Would there be any differences in the way the TL071 behaves compared to a TL072?

I have a circuit that is giving me (and others) trouble, and the only thing I have changed from the schematic is the opamp from a 1/2 used dual opamp to a single opamp...

[Somicide]

other than the obvious pinout changes, I can't think of why it would be different.  of course, my coffee still hasn't kicked in...

PnL,

Jeff

[spudulike]

Individual amp schematics are identical on the TL071/2/4 according to TI. TL071 includes offset null pins (1&5).

[Paul Marossy]

Individual amp schematics are identical on the TL071/2/4 according to TI. TL071 includes offset null pins (1&5).

That's what I thought. Hmm... this is weird. I must have done something wrong, but I can't find it for the life of me.

[Vsat]

The unused op amp in the dual should not be left unconnected with inputs floating or weird things might happen. Recommended method is to hook it up as a voltage follower (inv input wired to output) and connect the non-inv input to ground (if operated single-supply then connect it to the mid-supply reference or other voltage within the input common-mode range). This may or may not be the source of your problem..
Cheers, Mike

[Paul Marossy]

Vsat-

Hmm... that is a good point. Maybe in this particular circuit, the absence of that second "unused" opamp is a factor here.  :?

[Vsat]

Paul,
OK... so the original circuit actually has a TL072 with one-half unused.
The TL072 takes more current than the TL071 even if one half is not used. Maybe the problem is power-supply related... perhaps the circuit has a lot of series resistance in the supply wiring. I don't know what the circuit looks like, some op amp circuits deliberately put current limiting resistors in series with the op amp supply pins for various reasons.
Mike

[Paul Marossy]

Paul,
OK... so the original circuit actually has a TL072 with one-half unused.
The TL072 takes more current than the TL071 even if one half is not used. Maybe the problem is power-supply related... perhaps the circuit has a lot of series resistance in the supply wiring. I don't know what the circuit looks like, some op amp circuits deliberately put current limiting resistors in series with the op amp supply pins for various reasons.
Mike

In this case, there is V+ and V- of around 15V each and a bias control. Bias is determined by a 10K trimpot, end lugs connected to V+/V- and wiper providing the bias voltage to various points in the circuit.

EDIT: BTW, the mystery circuit is the Shaka Tube.  :wink:

[Vsat]

OK... looked at the Shaka circuit and peeked in the TI databook.

TI says <each> op amp in the TL072 (or the single op amp in the TL071)
typically draws 1.4 mA from the supply, and can draw as much as 2.5 mA each for extreme cases. And this is without a load on the output.

Looking at the schemo, there is a 470 ohm resistor in the (+)15V line and a 1K resistor in the (-)15V line. The (-)15V line has the most series resistance so let's look at that. Typical 1.4 mA current for  TL071 will produce a 1.4V drop across series 1K, reducing voltage at pin 4 of op amp to (-)13.6V. For TL072 taking 2x1.4 mA there would be 2.8V drop, resulting in (-)12.2V at pin 4. For worst case,  there will be (-)12.5V for TL071 and (-)10V for TL072.  Can be quite far away from fifteen volts then.

Also, the 1K series resistor in combination with the 470uF has a time constant of 0.47 seconds... will take a couple of seconds for the supply to come up to operating voltage. Non-inverting amps tend to be more susceptible to latch-up than inverting amps, I would check to see if the op amp has latched up.... output sticking to either supply rail.

A good test might be to jumper out the 1K and 470 ohm and see what happens.
Mike

[Vsat]

OK... looked at the Shaka circuit and peeked in the TI databook.

TI says <each> op amp in the TL072 (or the single op amp in the TL071)
typically draws 1.4 mA from the supply, and can draw as much as 2.5 mA each for extreme cases. And this is without a load on the output.

Looking at the schemo, there is a 470 ohm resistor in the (+)15V line and a 1K resistor in the (-)15V line. The (-)15V line has the most series resistance so let's look at that. Typical 1.4 mA current for  TL071 will produce a 1.4V drop across series 1K, reducing voltage at pin 4 of op amp to (-)13.6V. For TL072 taking 2x1.4 mA there would be 2.8V drop, resulting in (-)12.2V at pin 4. For worst case,  there will be (-)12.5V for TL071 and (-)10V for TL072.  Can be quite far away from fifteen volts then.

Also, the 1K series resistor in combination with the 470uF has a time constant of 0.47 seconds... will take a couple of seconds for the supply to come up to operating voltage. Non-inverting amps tend to be more susceptible to latch-up than inverting amps, I would check to see if the op amp has latched up.... output sticking to either supply rail.

A good test might be to jumper out the 1K and 470 ohm and see what happens.
Mike

[puretube]

http://diystompboxes.com/sboxforum/viewtopic.php?t=35100&postdays=0&postorder=asc&start=0

[MartyMart]

Vsat ... thats almost it :-) ( I'm one person having this problem )
Just jumpered both the 1k and the 470ohm connections
I can now get the signal to stay "on" much more, but it still sounds like
it's "popping" from rail to rail, in particular when "reducing" the gain control.
Still sounding quite "splatty" as well .....
I may have to try a more "muscle bound" PSU, see if that helps

Cheers,
Marty.

[puretube]

got it:
missing balancing resistors in the powersupply (470k each from both +&- to ground...)

"latchup"

schemo:
http://www.diystompboxes.com/pedals/schems/shakatube.gif
:?:

[Paul Marossy]

OK... looked at the Shaka circuit and peeked in the TI databook.

TI says <each> op amp in the TL072 (or the single op amp in the TL071)
typically draws 1.4 mA from the supply, and can draw as much as 2.5 mA each for extreme cases. And this is without a load on the output.

Looking at the schemo, there is a 470 ohm resistor in the (+)15V line and a 1K resistor in the (-)15V line. The (-)15V line has the most series resistance so let's look at that. Typical 1.4 mA current for  TL071 will produce a 1.4V drop across series 1K, reducing voltage at pin 4 of op amp to (-)13.6V. For TL072 taking 2x1.4 mA there would be 2.8V drop, resulting in (-)12.2V at pin 4. For worst case,  there will be (-)12.5V for TL071 and (-)10V for TL072.  Can be quite far away from fifteen volts then.

Also, the 1K series resistor in combination with the 470uF has a time constant of 0.47 seconds... will take a couple of seconds for the supply to come up to operating voltage. Non-inverting amps tend to be more susceptible to latch-up than inverting amps, I would check to see if the op amp has latched up.... output sticking to either supply rail.

A good test might be to jumper out the 1K and 470 ohm and see what happens.
Mike

This is very interesting. I hadn't thought of the possible implications by substituting a single opamp in place of a dual opamp. My reasoning was basically along the lines of "if a dual opamp works here, then a single opamp certainly ought to work". I guess I have learned this is not necessarily true.  :oops:

Ok then. In my build, I used a 13VAC 800mA power supply. Since my 13VAC 800mA wall wart puts out 15 volts when attached to this circuit, I used a 10ohm resistor in series with one leg of the power supply to get me 12.8VAC. These are my measured voltages with the circuit the way that I like it: On V+, 11.8V, V-, -14.6V and bias -13.6V. It seems to me that the power supply current available is also a factor in this mess...

[Paul Marossy]

got it:
missing balancing resistors in the powersupply (470k each from both +&- to ground...)

"latchup"

schemo:
http://www.diystompboxes.com/pedals/schems/shakatube.gif
:?:

Huh? I don't quite follow...  :?
Are you saying that with a TL071 this would be required?

[MartyMart]

got it:
missing balancing resistors in the powersupply (470k each from both +&- to ground...)

"latchup"

schemo:
http://www.diystompboxes.com/pedals/schems/shakatube.gif
:?:

Nope ... not it !
Just causes even more "splatty" performace than usual   :shock:
I'm going to re-make the board with a TL072 tomorrow and see what
happens ....  grrrrr

Marty.

[Paul Marossy]

Sorry for the headaches Marty.  
I can tell you that mine works fine with a TL072 in it.

[MartyMart]

Sorry for the headaches Marty.  
I can tell you that mine works fine with a TL072 in it.

Hey ... no problem .... I like to solve things  :-) and it helps to have guys
like you/Ton/Vsat around ......

I just took some voltage readings whilst playing with the circuit:

Input from the PSU "moves" its not a steady 16.5 volts, swings from
around 15 to 17.2
After the 680 Ohm resistor and at "board" entrance to tube etc its from
11.8 volts to 12.7 volts
So the reading at op-amp pin 7 corresponds : 23.6 volts to 25.4

This has no bearing on when the sound "craps out" ... this just happens
over a 1-2 minute cycle

I just got around 6 minutes of "working" circuit .... man this thing sounds
VERY nice.... so need to solve this.
New board and TL072 tomorrow.....

Marty, out for some dinner and a few beers now :-)

[Vsat]

Marty and Paul,
Since you are using an unregulated supply:

TI book says:  the absolute max rating for the TL071/TL072 supply pins is (+)18V and (-)18V... if this is exceeded even for a short duration the chip will almost certainly be damaged.  TI says that these measurements are made with respect to a reference voltage that is the midpoint between the two supplies. A good test would be to monitor the voltage between pin 4 and pin7 of the TL071, hopefully it never exceeds 2x15=30V under any conditions (even if the tube is unplugged, or the filament burns out).
Mike

[Paul Marossy]

Marty and Paul,
Since you are using an unregulated supply:

TI book says:  the absolute max rating for the TL071/TL072 supply pins is (+)18V and (-)18V... if this is exceeded even for a short duration the chip will almost certainly be damaged.  TI says that these measurements are made with respect to a reference voltage that is the midpoint between the two supplies. A good test would be to monitor the voltage between pin 4 and pin7 of the TL071, hopefully it never exceeds 2x15=30V under any conditions (even if the tube is unplugged, or the filament burns out).
Mike

Yep, I am aware of that one.  :wink: