TL072 vs. TL082... sound and noisefloor

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: Kent, Joe., luketeaford, lisa

ttown23
is banned
is banned
Posts: 1157
Joined: Fri Oct 12, 2012 5:20 pm
Location: Los Angeles, CA

TL072 vs. TL082... sound and noisefloor

Post by ttown23 » Mon Jan 20, 2014 8:37 pm

Just wondering if I can replace TL072 with TL082.... well, OK, I know that technically they are interchangable... but interested to hear opinions on which is the 'better' OP-AMP ;)
I'm banned for dishonest and unethical behaviour.

User avatar
widdly
Super Deluxe Wiggler
Posts: 1096
Joined: Wed Jan 21, 2009 1:31 am
Location: australia

Post by widdly » Mon Jan 20, 2014 8:57 pm

Wouldn't it depend on the circuit? Maybe it's a comparator on the input to a clocking module, maybe it is buffering a CV, maybe it is the front end of a microphone preamp, maybe there is some other noisey or distorting part that will hide the marginal difference between the two parts.

ttown23
is banned
is banned
Posts: 1157
Joined: Fri Oct 12, 2012 5:20 pm
Location: Los Angeles, CA

Post by ttown23 » Mon Jan 20, 2014 9:10 pm

widdly wrote:Wouldn't it depend on the circuit? Maybe it's a comparator on the input to a clocking module, maybe it is buffering a CV, maybe it is the front end of a microphone preamp, maybe there is some other noisey or distorting part that will hide the marginal difference between the two parts.
Yes, it will... I was looking for overall impressions though, all other things being equal... :tu:

Mostly thinking about it being used in front of a preamp circuit... how the overall audio quality would be with 72 vs 82.

Strangely, there isn't much info. on this out there, and the TL072 is now an obsolete part...
I'm banned for dishonest and unethical behaviour.

slow_riot
Super Deluxe Wiggler
Posts: 4002
Joined: Tue Jun 30, 2009 8:58 pm
Location: uk

Post by slow_riot » Mon Jan 20, 2014 9:26 pm

i don't think you can easily tell the impact that a single opamp makes in the signal chain through listening, certainly not the difference between 082 and 072. Perhaps if you have a lot of experience and know exactly what you are looking for. Maybe if you're using the opamp wrong you could hear it (e.g. lots of people complain about the OP275 because they dropped it into a TL0x circuit and expected it to work perfectly.)

I think the best thing you can do is look at the specs of the opamp, and compare that to the requirements of the circuit you are using it in, and your budget.

User avatar
Nantonos
Super Deluxe Wiggler
Posts: 1275
Joined: Sun Mar 03, 2013 1:00 pm
Location: Boston, MA

Post by Nantonos » Mon Jan 20, 2014 9:28 pm

Why do you say TL072 is obsolete? Its widely available and has no obsoletion notices at any of the suppliers I use.

The parameters of 082 vs 072 in the datasheets look very similar (082 has lower input bias current) and the graphs of equivalent input noise vs. frequency, and of THD vs. frequency, look near identical for the two devices.

If you are using it in an audio preamp, wouldn't an NE5532, or LM4562, or LME49720 be a better choice?
ondes | current rack
I am afraid a firmware change will not be able to turn a rather expensive 16-bit DAC into a 16-bit ADC, and flip all those op-amps :)

ttown23
is banned
is banned
Posts: 1157
Joined: Fri Oct 12, 2012 5:20 pm
Location: Los Angeles, CA

Post by ttown23 » Mon Jan 20, 2014 9:44 pm

[quote="Nantonos"]Why do you say TL072 is obsolete? Its widely available and has no obsoletion notices at any of the suppliers I use.
[quote]

It's obsolete at mouser.com...!
I'm banned for dishonest and unethical behaviour.

User avatar
fluxmonkey
Super Deluxe Wiggler
Posts: 1850
Joined: Sat Oct 18, 2008 10:05 pm
Contact:

Post by fluxmonkey » Mon Jan 20, 2014 9:54 pm

595-TL072CP, "5,843 Can Ship Immediately"... and that's just 1 of 10 thru-hole listed...
www.fluxmonkey.com

BUILDER FOR HIRE - kits or custom, email me with your needs

ttown23
is banned
is banned
Posts: 1157
Joined: Fri Oct 12, 2012 5:20 pm
Location: Los Angeles, CA

Post by ttown23 » Tue Jan 21, 2014 1:23 am

fluxmonkey wrote:595-TL072CP, "5,843 Can Ship Immediately"... and that's just 1 of 10 thru-hole listed...
Well that's cool. Thanks. I must have been ordering an old part number; it came up obsolete!

But, back to the topic at hand... comparisons between TL072 and TL082. Someone must have done a side-by-side?
I'm banned for dishonest and unethical behaviour.

User avatar
Synthbuilder
Super Deluxe Wiggler
Posts: 3060
Joined: Tue Mar 24, 2009 9:12 am
Location: Cumbria, UK

Post by Synthbuilder » Tue Jan 21, 2014 3:51 am

Years ago... there was a difference. But this is all from memory because I sure don't have the databooks to back it up. The TL072 was identical to the TL082 except that the TL072 had lower noise and the TL082 had lower offset voltage. I guess it was the same die and then parts were selected by testing.

Tony

User avatar
sonicwarrior
Super Deluxe Wiggler
Posts: 3815
Joined: Tue Sep 29, 2009 3:36 pm
Location: Cologne, Germany

Post by sonicwarrior » Tue Jan 21, 2014 5:05 am

While we are at it:
Synthbuilder wrote:The TL072 was identical to the TL082 except that the TL072 had lower noise and the TL082 had lower offset voltage.
So the TL072A has both lower noise and the lower offset?

The MOTM-440 BOM listed TL072A but I put TL072CP in there because I didn't have them, some Oakley module lists the A version in the BOM so I'm about to order them.

Also is there a difference between TL072ACPE4 and TL072ACP? The TL072ACPE4 are in stock @ Mouser while the TL072ACP are not which is why I'm asking. Couldn't find anything about the E4 suffix in the TI datasheet.
Soundcloud
For sale: Jürgen Haible Tau Phaser (MOTM format)

User avatar
guest
Super Deluxe Wiggler
Posts: 4927
Joined: Mon Aug 19, 2013 11:49 am

Post by guest » Tue Jan 21, 2014 5:31 am

i remember being very confused about these 2 opamps a number of years ago, and combing both datasheets for any difference. the only thing i remember finding was slightly lower noise on the 072, and the THD knee was at 20kHz versus 10kHz for the 082.

after a quick double check, this seems to be the case. the bias current looks a bit higher, and the GBW a bit lower on the 072, so im guessing they turned the bias current up on the input stage of the 072 to give better noise and distortion characterstics for audio applications, at the expense of input bias current and GBW. since they seem to be equally available, i would go with the 072 for audio stuff, although the differences will most likely be subtle.

and as a side note, the 082 datasheet is amazing, so many good application circuits in it (at least the TI one). i learned a ton from it when i was starting out.
openmusiclabs.com

User avatar
cloudscapes
Wiggling with Experience
Posts: 388
Joined: Mon Jan 23, 2012 6:01 pm
Location: montreal

Post by cloudscapes » Tue Jan 21, 2014 9:42 am

072/074 is as likely to become obsolete as the 555 at this point ;)

OPA4134 is a nice hifi drop-in replacement to the 074. bit pricer though, and there may be other noise bottlenecks in a circuit other than opamps

User avatar
daverj
Vintage Video Wiggler
Posts: 8514
Joined: Fri Jun 19, 2009 11:09 am

Post by daverj » Tue Jan 21, 2014 11:51 am

The TL07x series is certainly not obsolete. Take the letters off the end and do another search and you'll find lots of them. Certain versions become obsolete as they get replaced by a different version. This happened mostly as the versions containing lead were replaced with ones that don't.

The TL07x was released as an improved version of the 8x series. Since then they have improved the 8x to the point where there is very little difference between them. There are a couple of specs in the standard 8x that are very slightly worse than the 7x, and one that is slightly better in the "typical" value but worse in the maximum value.

The improved versions of each, the "A" and "B" versions, are even closer to being the same.

The 7x comes in some additional package versions (all SMT) that the 8x doesn't come in.

If they ever discontinue one of them, my guess is it will be the 8x.

User avatar
daverj
Vintage Video Wiggler
Posts: 8514
Joined: Fri Jun 19, 2009 11:09 am

Post by daverj » Tue Jan 21, 2014 11:56 am

After posting above I noticed the thread title and forgot to mention:

The noise figures, slew rate, and THD are identical for both series.

ttown23
is banned
is banned
Posts: 1157
Joined: Fri Oct 12, 2012 5:20 pm
Location: Los Angeles, CA

Post by ttown23 » Tue Jan 21, 2014 12:15 pm

Really useful information- thanks guys!

I think I will go with the TL072... and if I can get my hands on them, the 'A' or 'B' versions. I had always used the TL072CP variety with good success... so based on one of the posts above, is there really any difference between TL072'A' or 'B' and the 'CP' variety?
I'm banned for dishonest and unethical behaviour.

User avatar
daverj
Vintage Video Wiggler
Posts: 8514
Joined: Fri Jun 19, 2009 11:09 am

Post by daverj » Tue Jan 21, 2014 12:27 pm

The "C" is the temperature range (C = commercial, I = industrial), and the "P is the package type (DIP). The improved version letters (A and B) come between the numbers and those letters. So the improved versions of the TL072CP are the TL072ACP and TL072BCP.

The main difference with the A and B versions is lower input offset voltage. A few other specs have slightly better "minimum" specs while the "typical" values stay the same.

The main difference with better input offset voltages is that DC coupled circuits, especially ones with a lot of gain, are more likely to work exactly the same if you make several copies. Those voltages are very small in all versions though, so in many circuits you would never notice the difference.

ttown23
is banned
is banned
Posts: 1157
Joined: Fri Oct 12, 2012 5:20 pm
Location: Los Angeles, CA

Post by ttown23 » Tue Jan 21, 2014 1:02 pm

daverj wrote:The "C" is the temperature range (C = commercial, I = industrial), and the "P is the package type (DIP). The improved version letters (A and B) come between the numbers and those letters. So the improved versions of the TL072CP are the TL072ACP and TL072BCP.

The main difference with the A and B versions is lower input offset voltage. A few other specs have slightly better "minimum" specs while the "typical" values stay the same.

The main difference with better input offset voltages is that DC coupled circuits, especially ones with a lot of gain, are more likely to work exactly the same if you make several copies. Those voltages are very small in all versions though, so in many circuits you would never notice the difference.
Got it. This sounds like a case of 'diminishing marginal returns' with the 'A' and 'B' versions then...

Probably I'll go ahead and stick with the trusty 'C' for audio applications then ;)
I'm banned for dishonest and unethical behaviour.

User avatar
Nantonos
Super Deluxe Wiggler
Posts: 1275
Joined: Sun Mar 03, 2013 1:00 pm
Location: Boston, MA

Post by Nantonos » Wed Jan 22, 2014 2:58 am

Synthbuilder wrote:Years ago... there was a difference. But this is all from memory because I sure don't have the databooks to back it up. The TL072 was identical to the TL082 except that the TL072 had lower noise and the TL082 had lower offset voltage. I guess it was the same die and then parts were selected by testing.
Well, if you look at the datasheets (come on, its not hard), the 072 and 082 have identical noise performance and the same input offset voltage (there is more difference between the three grades of each than there is between 072 and 082).

I have seen this myth about the noise being different several times, wonder where it comes from.
ondes | current rack
I am afraid a firmware change will not be able to turn a rather expensive 16-bit DAC into a 16-bit ADC, and flip all those op-amps :)

User avatar
Synthbuilder
Super Deluxe Wiggler
Posts: 3060
Joined: Tue Mar 24, 2009 9:12 am
Location: Cumbria, UK

Post by Synthbuilder » Wed Jan 22, 2014 4:16 am

Nantonos wrote:Well, if you look at the datasheets (come on, its not hard), the 072 and 082 have identical noise performance and the same input offset voltage.
They do now. What I was saying is that thirty-five years ago when they were introduced I remember them not being the same.

Indeed, if they were the same, why have two part numbers?

Tony

User avatar
Synthbuilder
Super Deluxe Wiggler
Posts: 3060
Joined: Tue Mar 24, 2009 9:12 am
Location: Cumbria, UK

Post by Synthbuilder » Wed Jan 22, 2014 4:30 am

Well, actually it was quite hard. I had to go into the loft to find this one.

From Texas Instruments' "The Bi-FET Design Manual" from 1984. This clip is from the introduction to the devices at the start of the booklet:

Image

The actual data shows typical noise voltage of 25nV/rt Hz for the TL08X and 18nV/rt Hz for the TL07X.

Tony

User avatar
Plop
Veteran Wiggler
Posts: 505
Joined: Sat Mar 06, 2010 10:15 am
Location: Utrecht, NL

Post by Plop » Wed Jan 22, 2014 3:04 pm

Nantonos wrote:I have seen this myth about the noise being different several times, wonder where it comes from.
I don't know if it's obsolete information but TI still states on their website that the noise floor of the TL07x series is lower than the TL08x.

Image

Wayfinder
1-Post Wiggler
Posts: 1
Joined: Wed Apr 23, 2014 10:35 am

Post by Wayfinder » Wed Apr 23, 2014 10:42 am

Nantonos wrote:If you are using it in an audio preamp, wouldn't an NE5532, or LM4562, or LME49720 be a better choice?
Greetings Nantonos. Would be interested in reading why you prefer the NE5532 or LM4562. Stuff like that is always good to know. Thanks.

User avatar
Nordcore
Veteran Wiggler
Posts: 583
Joined: Fri May 10, 2013 2:05 pm
Location: Germany

Post by Nordcore » Wed Apr 23, 2014 2:09 pm

Nantonos wrote:If you are using it in an audio preamp, wouldn't an NE5532, or LM4562, or LME49720 be a better choice?
Depends on the source resistance.
The input noise current of the bipolar types is in the 2pA/SQR(Hz) range - which gives 18nV extra input noise at 10kOhms source resistance.

User avatar
Nantonos
Super Deluxe Wiggler
Posts: 1275
Joined: Sun Mar 03, 2013 1:00 pm
Location: Boston, MA

Post by Nantonos » Wed Apr 23, 2014 3:00 pm

Synthbuilder I missed your earlier reply, sorry. Thanks for digging out the 30 year old datasheets. As you say, when introduced they had somewhat different noise specs of 25nv/√Hz vs. 18nv/√Hz. Now both are specced at the latter value. The noise vs. frequency plots are very similar, too.

Plop, I guess marketing slogans have greater longevity than technical specifications do.

Nordcore, yes indeed. Note that my original statement was not "I prefer NE5532 over TL072 for everything" but instead "If you are using it in an audio preamp, wouldn't an NE5532, or LM4562, or LME49720 be a better choice?". An audio preamp typically needs low noise and high gain, and typically doesn't care about current draw. So devices with low voltage noise, higher current noise, and much lower resistor values than are typical in modular synths (for low Johnson noise) can give significantly better results. Which is why commercial pro audio is full of low cost NE5532 opamps and not TL072. The fact that 5532 is only stabilised for a closed loop gain of 3x, and has much lower input impedance than JFET designs, is not a concern in that application.

I'm recalling some noise vs. frequency and distortion vs. frequency graphs for TL072, NE5532 and LM4562 but can't place where I saw them. Will post again if I remember where they were.

For noise-sensitive synth applications at higher impedances, the LME49720 low voltage noise of 2.7nV/√Hz is much better than TL072 because 18nV/√Hz may have been low noise 30 years ago but isn't really now. Of course, the contribution of the surrounding circuitry still needs to be considered.
ondes | current rack
I am afraid a firmware change will not be able to turn a rather expensive 16-bit DAC into a 16-bit ADC, and flip all those op-amps :)

User avatar
Nordcore
Veteran Wiggler
Posts: 583
Joined: Fri May 10, 2013 2:05 pm
Location: Germany

Post by Nordcore » Wed Apr 23, 2014 5:39 pm

Nantonos wrote:The fact that 5532 is only stabilised for a closed loop gain of 3x,
NE5534 is stable >=3 (and has compensation pins), NE5532 is unity gain stable.
Nantonos wrote:For noise-sensitive synth applications at higher impedances, the LME49720 low voltage noise of 2.7nV/√Hz is much better than TL072
No.
At higher impedance the pretty high current noise of the LME49720 will become dominant - so the LME49720 will be noisier than the TL072.
(Roughly estimated: if the inputs sees more than 2kOhm the LME49720 should be replaced by something better fitting ... )

Lets assume a 20kOhm resistor at the input:
Johnson noise: 18nV//√Hz
Input voltage noise TL072: 18nV//√Hz
total: 25nV//√Hz

Input noise current LME49720: 1,6pA/√Hz *20kOhm= 32nV/√Hz
Johnson noise: 18nV//√Hz
total: 36nV/√Hz

Going cheap but good, I'd select a LF356, JFET input with 11nV//√Hz, total 21nV/√Hz.

Noise figures:
LF356: 1,4dB
TL072: 3dB
LME49720: 6dB

Post Reply

Return to “Music Tech DIY”