Texas Instruments TL2072 instead of TL072

[jetam]

It has broader range of supply voltages and all other specs are better therefore it should work. Just make sure that the pin configuration is the same. I believe that it is, but it never hurts to doublecheck.

[znus]

Originally Posted by jetam

It has broader range of supply voltages and all other specs are better therefore it should work. Just make sure that the pin configuration is the same. I believe that it is, but it never hurts to doublecheck.

Thanks! Pin configutration are the same between the two

http://www.ti.com/lit/ds/symlink/tl072.pdf

http://www.ti.com/lit/ds/symlink/tle2072.pdf

Shame the 2072 is ~6 times the price

[guze]

You probably don't need to change all, some maybe are just buffers. In that picture it looks like they are in the eq.
Buy some and try with the other ones that you have. Neither is super fast and should not cause problems/oscillations.
Try a channel first, and see/listen if it's worth it. You will notice some difference, but don't expect a big difference.

[brianroth]

LF-353N should be a drop-in replacement, marginally better in some respects than a TL072.

Bri

[znus]

@guze. Definitely. I'll try with 1-2 channels before going all in. I'm not expecting to actually hear any difference and if I do - that would be the big thing of 2012 for me!

@brianroth. Aha! Thanks! And at a significally lower price too. Sweet

[brianroth]

Try them and see if you notice any difference. The LF353's were National's alternative to TI's TL072. Some desk manufacturers preferred them Back In The Day. The "N" version is the 'better" version, iIRC.

Dunno what to think now that TI bought-out National....

Bri

[DrSound]

The primary used TL072 has 18nV/SQRT(Hz) noise density.
Offered TLE2072 has 17nV/SQRT(Hz) – practically the same.
For this reason it not makes sense to perform such replacement.
It makes sense to use this opamp only for new designs.
I personally recommend to use the LME49880, which has
only 7nV/SQRT(Hz) noise density.

[znus]

@bri. I think the fact that you get three of the 353's for one 2072 is reason enough

@DrSound. 49880 looks interesting. My local dealer doesn't carry it but DigiKey have them.

LME49880MR/NOPB National Semiconductor | LME49880MR/NOPB-ND | DigiKey

Edit: Hmm, the 49880 in the link are surface mount and i need through hole.

[doulos30]

you will hear zero improvement from using a 2072. I bought 50 of them at 2.50 cent's a piece there's not enough improvement between them in fact there's no improvement at all total waste of money

[Jim Williams]

Got scope?

[living sounds]

I think they sound pretty different. Try it. If you need more TL2072 I could sell you some, I bought a lot of them for upgrading several pieces of gear I own and still have some left.

[znus]

@doulos. I might try 2072 on channel 1 & 2 and compare it with TL072 3 & 4, feeding the two stereopairs the same signal.

@Jim. Nope, but currently eyeballing one. Then I just gotta figure out how to use it It was 20 years since I last fiddeled around with a scope (in high school). My wife will love me adding a scope to my collection of gear...


@livingsounds. We could definitely work that out. They are $3 a pop in Sweden

[znus]

Did a little test just now. I changed one TL072 on channel 1 (the IC2 one) and replaced it with a 2134 and something definitely audible happened.

If I look at the master bargraph boosting +15 db MF 8 kHz on channel 1 and on channel 2 (channel 2 = all TL072) I get 1-2 db higher peaks on Channel 2 but it sounds a lot grainier. None of the channels are peaking so it's not distortion.

Maybe the Q is wider och channel 2? Maybe I moved up the freq's on channel 1 with the 2134? SOMETHING is different though and channel 1 is easier on the ears. Maybe it's the db difference playing me a trick but 2134 is, without a doubt, changing something for the better on channel 1. At least when boosting the eq. Less grain

Wierd, and interesting. I did not expect to hear a difference, so it's not me wanting to hear it.

Hmmmm What to do what to do....

Edit: Something definitely happened with freq's (or Q). Listening to samples of a ride cymbal I hear more of the bell on the TL072 channel 2 and more sizzle on the BB 2134 channel 1

[Karloff70]

Originally Posted by znus

Did a little test just now. I changed one TL072 on channel 1 (the IC2 one) and replaced it with a 2134 and something definitely audible happened.

If I look at the master bargraph boosting +15 db MF 8 kHz on channel 1 and on channel 2 (channel 2 = all TL072) I get 1-2 db higher peaks on Channel 2 but it sounds a lot grainier. None of the channels are peaking so it's not distortion.

Maybe the Q is wider och channel 2? Maybe I moved up the freq's on channel 1 with the 2134? SOMETHING is different though and channel 1 is easier on the ears. Maybe it's the db difference playing me a trick but 2134 is, without a doubt, changing something for the better on channel 1. At least when boosting the eq. Less grain

Wierd, and interesting. I did not expect to hear a difference, so it's not me wanting to hear it.

Hmmmm What to do what to do....

Stay with the 2134 in there for a couple of days at least before you decide what it does is a good thing and replace everything with them. The chocolatey slickness can become a bit like eating a whole slippery chocolate cake. Loss of paperyness. YMMV

[znus]

Originally Posted by Karloff70

Stay with the 2134 in there for a couple of days at least before you decide what it does is a good thing and replace everything with them. The chocolatey slickness can become a bit like eating a whole slippery chocolate cake. Loss of paperyness. YMMV

Aye sir!

Just did IC2 on channel 2 and the same thing happened. Not that it was a surprise but I had to rule out other possibilites, like wierd behaving TL072's and what not.

I guess you didn't see my last edit, about the ride cymbal bell thing (since it's not included in your quote)?

Will be interesting to change the IC1 and 3 with 2134's and/or 2072's. I wish I knew what the IC1, 2 & 3 did. Are they HF, MF and LF or are IC1 the amp, the IC2 the eq and IC3 for summing? We will see (hear)!

[Karloff70]

I tried those in my desk eq, which is designed for TL072's, but actually has 5532's in there.....and I went back to the 5532's. But as everyone will tell you, the circuit and the way it makes use of the chip will dictate what difference you get. Mine wasn't to my liking, although initially enticing.

[znus]

My little desk has a bunch 5532's on the master. Looks like they are for sends, while the returns have 072's. That's just me guessing though.

Ok, next up is a 2134 on IC1. If HF changes just as noticable we can rule out it's for line and mic amplification. If nothing happens frequencywise and I gotta trim up or down - it's amp.

Back in 15

[living sounds]

I think for EQ (especially those designed for TL072) the OPA2134 (or even better OPA2132) rules. I went through lots of op amps and did whole mixes with sets of them, and the OPA2132(4) was what I kept in my console in the EQ section. In the signal path it sounded kinda boring and bland though.

[znus]

Back!

HF sounds less honky on the 2134 channel. Cant decide if its frequency band or q-width? Gotta try the LF, which should be easier to hear if the whole frequency spectrum has been moved up.

@living sounds. Aha! Then I gotta atleast try a set of 2132's

This is fun! Never thought I'd hear any differencies and I'm usually very suspicious about modifications, since a lot of them doesn't even show any change or improvement on scientific tests. It's one thing doing all out Jim Williams-modifications but changing a simple opamp shouldn't have this impact.

Ok, time for IC3!

[living sounds]

I always match the frequency response with white noise graphically and via reversed phase cancellation. It doesn't show up in static THD measurements either, this seems to be a dynamic response difference between op amps.

[JohnRoberts]

I've used many TL07x opamps in console paths but typically at modest low closed loop gain and driving relatively high impedance paths. In such applications their transfer function and therefore "sound" should be dominated by the negative feedback components (value and behavior).

The more difficult the circuit application, like large bus summing amps, high gain stages, or output drivers, the more chance of some slight audible difference.

Note: many high gain mic preamps using TL07x incorporate a circuit topology where discrete input devices deliver both the low input noise and high open loop gain, so once again there are rather small errors available to improve upon by throwing expensive newer opamps at the design.

I do not argue with people about what they say they hear, but IMO a lot of this is a waste of time and money. If you are serious, start making some actual measurements and search out any performance parameters that may have deteriorated over the years, or leave room for targeted improvement with empirical cause and effect justification.

This is science not magic...

JR

[znus]

And there goes IC3.

LF - can't say I'm hearing the same drastic change as I did on the MF. I think I can rule out I'm changing the frequencies since it's definitely working in the same area. Well, I'm not saying I can hear the difference between 60 Hz and 66 Hz, but still... ..

living & JR. I wish I had the knowledge and gear to measure stuff but for now I just gotta go with what I can hear. I do not believe in magic. As I wrote earlier "I'm usually very suspicious about modifications, since a lot of them doesn't even show any change or improvement on scientific tests.".

Since it's only 3 IC's per channel * 8 it's not that much time & money. It's fun and educational and I'm gonna change all of them + beef up the PSU.

[Jim Williams]

I ended up pulling out my decades old stock of TLE2072's as I used one in an Otari 12 deck. The inverting pins on both opamps connect together on the output balanced drive circuit. National LME's don't like that. Since the trace connects them below the socket, re-wiring with some snubber resistors isn't an option. The TLE works fine there and was chosen because it will work at + - 20 volts, not many will. I didn't bother with the BB OPA2604, I think that chip sounds like crap.

The best mod was ripping out that Harris SIP preamp chip they used for the playback head amp. I used a "surfboard" adaptor with a LME49990 opamp at .85 nv noise, rewired the design, changed the power distribution and added 1000 uf psu local bypass caps and removed the input el cap off the head, now direct coupled with a ferrite bead for rf trapping.

What a noise drop! Slew rate went up about 20 times too, the low end is solid. Fog is removed.

[znus]

Jim, since you are very familiar with Soundcraft - are there any mods one can do to the CPS150? I'm guessing my 8 channel desk isn't that power hungry but since I'm changing 072's (3 mA) to 2134's (8 mA) the current will go up. 30 opamps will demand 150 mA more.

[living sounds]

Originally Posted by znus

Jim, since you are very familiar with Soundcraft - are there any mods one can do to the CPS150? I'm guessing my 8 channel desk isn't that power hungry but since I'm changing 072's (3 mA) to 2134's (8 mA) the current will go up. 30 opamps will demand 150 mA more.

Do a search, the CPS150 mods have been covered here.

[Jim Williams]

I use 15,000 uf Nichicon VZ mains caps, 470 uf secondary caps, replace the 1/4 watt 2.7k phantom divider resistor with 1/2 watt (it's the burned one) and check the rectifier bridge.

You may need a bigger supply.

[znus]

Hmmm...just registered an account on DigiKey. Maybe OPA2132 is the way to go since they only draw 4 mA

(and still beefing up the CPS150 a la Jim Williams reciept)

edit. The TI 2134 also draws 4 mA while the BB OPA2134 draws 8 mA

[living sounds]

Originally Posted by znus

edit. The TI 2134 also draws 4 mA while the BB OPA2134 draws 8 mA

Same thing. TI bought Burr Brown a long time ago. You probably mixed up current draw per chip vs. current draw per amp (there are two amps in a dual).