TLO-72CN vs TLO-72CP

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

AusTex64

Well-known member
Joined
Jun 3, 2013
Messages
525
It appears TLO-72CN is obsolete. I can't find them new from Mouser or anyone else.

I did find TLO-72CP on Mouser. The main difference I see is a slightly slower slew rate - 16V/us for TLO-72CN vs. 13V/us for TLO-72CP.

The application is in the protection circuit in an Ampeg SVT Classic. Here's the schematic: http://bustedgear.com/images/schematics/Ampeg_SVT-CL_fault.pdf I'm thinking the reduced slew rate shouldn't make much difference in that application. What say you?

Thanks as always for the help.
 
I am not aware of any such slew rate difference...  The typical tl07x was always 13 V/uSec. I could imagine a slight difference from foundry improvements over the decades since this part was created.

I suspect they are phasing out through hole packages as demand tapers off...

The A,B,C IIRc was about a relaxed input voltage offset spec,

the N and P was package.

If you can get the TL072 anything in a package you can use it will probably be OK... 

Glancing at the schematic it looks like they are using them as comparators (they are op amps not comparators). While not true comparators, bifet opamps have very low input bias current so likely OK.

JR
 
TL072CN is (was) a STMicroelectronics part with a 16V/us SR and end of life anounced some years ago.
TL072CP is a TexasInstruments part with 13V/us SR.
Letter "C" is temperatur range 0-70° C.
There is no letter "O" in the part number.
 
I seriously doubt slew rate would matter for just about anything audio. I get the impression that TLV2172 is sort of the next revision of that part and it has a slew rate of 10V/us. Slew rate matters if you're doing something like buffering digital video and then you use something with a slew rate of 1000V/us.
 
ruffrecords said:
Didn't somebody do a lower noise version of the TL072?

Cheers

Ian

The TL072 is the lower noise version of the TL082.

There's the more recent TLE2072, much lower noise floor and higher slew rate, often suggested as an upgrade to the TL072 (including by the manufacturer). But it requires much more care with regards to decoupling for stability. I never liked the sound in any of the circuits I tried it, but it seems to work well in my Vermona Perfourmer MK2 synth...
 
ruffrecords said:
Didn't somebody do a lower noise version of the TL072?

Cheers

Ian
Yup, the TL07x is about 40 years old so process technology has improved since then. That said the high (fet) input impedance, relatively low noise (3uV), and more than adequate slew rate (13v/uSec)made these parts better than needed for most serious line level audio paths. Another old soldier about the same age the NE553x was likewise better than fast enough for audio and quiet enough for line level use.  I have used truckloads of both over the decades.

For today's TMI I am working on a personal project, where I need a simple 6 channel bump box to get the typical 1V AC output from my cheap surround decoder up to 2.5V AC  to drive my Hypex amp modules to full scale output.  Ideally I want to run the bump box from the same 5V supply to KISS. After a quick search here are some modern CMOS op amps that don't suck. I found one with great specs that will go rail to rail from 5V supply, but it is a little pricey, I found another dual CMOS op amp that is more than good enough for around $0.80.... 

So long story shorter, op amp technology has continued improving over the decades since, while the old soldiers are still more than adequate (for slow moving audio so don't expect much audible improvement). After 4 decades it is not that surprising to see some go end of life. Don't worry the IC makers will have multiple alternate substitutes to sell you if you don't mind the price  increase.

JR 
 
John, the LM4562 is now down to a price that it makes little sense to use anything else unless you are dealing with high impedances. The LM4562 has lower noise, lower THD, higher slew rate, higher bandwidth and lower offset  than the good old 553x, it does have, however, higher input noise current than the 553x which makes it less suitable for things like phono amps. It seems like the 4562 is the new audio workhorse, I personally prefer the OPA1612 but it is 6 to 7 times more expensive than the 4562 and its only available in SMT.

The 4562 seems to have at some point some problems with noise, it was attributed to the fabrication process and seems to have been corrected since.

In my opinion, the OPA2134 is the modern  replacement for the TL072 when it comes to audio, it is considerably quieter than the TL072 (18nV/rtHz VS 8nV/rtHz), lower distortion, higher bandwidth and slew rate, lower offset, but it is around 5 times more expensive than the TL072.

The ultimate audiophile FET opamp is the NJR MUSES03, it only comes in a single version and costs the modest amount of $59 USD at Mouser. I've seen people replacing OPA627 with NJRs. The MUSES02 is the bipolar version, costing $46, with its unimpressive 5 V/usec slew rate, 4.5 nV/rtHz and 5.8MHz unity gain bandwidth frequency, those specs are considerably worse than the 4562 but the manufacturer states that it is "optimized for high-end audio and professional audio applications with advanced circuitry and layout, unique material and assembled technology by skilled-craftwork. " what ever that means...
 
The factors that determine replacement compatibility are offset, bandwidth and supply. If the circuit is not designed for offset, you could get scratchy pots or distortion or other weird behavior. If the bandwidth is higher, it usually requires bypass caps at the supply pins so if an existing PCB doesn't have those, you're out of luck. If the supply does provide enough current, you can get noise, oscillation and distortion. TL072 is low offset, low bandwidth and low supply. The closest match for that that TI seems to be pushing is TLV2172 which has actually been around for a long time as one of the "SoundPlus" chips AKA OPA1688. Many of the other chips mentioned do not comply with that. If you're replacing a bunch of amps, the difference between 4mA/amp and 1.5mA/amp is not insignificant. And incidentally TLV2172 has 9nV/rtHz noise, 0.0002% THD if that sort of thing really matters to you.
 
squarewave said:
The factors that determine replacement compatibility are offset, bandwidth and supply. If the circuit is not designed for offset, you could get scratchy pots or distortion or other weird behavior. If the bandwidth is higher, it usually requires bypass caps at the supply pins so if an existing PCB doesn't have those, you're out of luck. If the supply does provide enough current, you can get noise, oscillation and distortion. TL072 is low offset, low bandwidth and low supply. The closest match for that that TI seems to be pushing is TLV2172 which has actually been around for a long time as one of the "SoundPlus" chips AKA OPA1688. Many of the other chips mentioned do not comply with that. If you're replacing a bunch of amps, the difference between 4mA/amp and 1.5mA/amp is not insignificant. And incidentally TLV2172 has 9nV/rtHz noise, 0.0002% THD if that sort of thing really matters to you.

You are being a bit dramatic, I'm not saying you are not right, but he is not replacing the opamps of a 32 channel console, if you replace a couple chips and they eat 8mA more current in total, there is no problem.
 
user 37518 said:
You are being a bit dramatic, I'm not saying you are not right, but he is not replacing the opamps of a 32 channel console, if you replace a couple chips and they eat 8mA more current in total, there is no problem.
It might not be a problem. But it so happens that in this particular case it most definitely would be. There are 14 amps in that schem at 1.4mA. Using OPA2134 @ 4mA each is a difference of 36mA. That's not insignificant. And there are no bypass caps but all of the amps you mentioned would need them.
 
squarewave said:
It might not be a problem. But it so happens that in this particular case it most definitely would be. There are 14 amps in that schem at 1.4mA. Using OPA2134 @ 4mA each is a difference of 36mA. That's not insignificant. And there are no bypass caps but all of the amps you mentioned would need them.

I can get 36 mA from a lemmon,  we are talking 2 white LEDs, I dont think the PSU is going to faint from that. On the other hand, the lack of bypass caps does make a difference in which case is the TLV2172 you recommended also going to work without those?
 
user 37518 said:
John, the LM4562 is now down to a price that it makes little sense to use anything else unless you are dealing with high impedances. The LM4562 has lower noise, lower THD, higher slew rate, higher bandwidth and lower offset  than the good old 553x, it does have, however, higher input noise current than the 553x which makes it less suitable for things like phono amps. It seems like the 4562 is the new audio workhorse, I personally prefer the OPA1612 but it is 6 to 7 times more expensive than the 4562 and its only available in SMT.

The 4562 seems to have at some point some problems with noise, it was attributed to the fabrication process and seems to have been corrected since.
The 4562 is nice at higher rail voltage but the distortion plots for +/-2.5V start to rise well before 1 VAC. A larger concern is 10-12 mA per (package or per amplifier?), of course less current from 5V rail. 
In my opinion, the OPA2134 is the modern  replacement for the TL072 when it comes to audio, it is considerably quieter than the TL072 (18nV/rtHz VS 8nV/rtHz), lower distortion, higher bandwidth and slew rate, lower offset, but it is around 5 times more expensive than the TL072.
The OPA2134 was the really good op amp I mentioned but looks marginal drive capability for rail to rail use at 5V rail

I haven't bought TL07x in small quantity ever, but even digikey has them for $0.41 so more like 10x but academic because tl072 is not happy for 5V rail.
The ultimate audiophile FET opamp is the NJR MUSES03, it only comes in a single version and costs the modest amount of $59 USD at Mouser. I've seen people replacing OPA627 with NJRs. The MUSES02 is the bipolar version, costing $46, with its unimpressive 5 V/usec slew rate, 4.5 nV/rtHz and 5.8MHz unity gain bandwidth frequency, those specs are considerably worse than the 4562 but the manufacturer states that it is "optimized for high-end audio and professional audio applications with advanced circuitry and layout, unique material and assembled technology by skilled-craftwork. " what ever that means...
I have spent the last few hours search through data sheets fo a good solution.

Another variable that I did not mention is that the hypex class D amp modules have a relatively low input impedance (1.8k) , so not only do I want rail to rail from 5V but I want to drive < 1.8k.

While I wasn't looking there are now a handful of CMOS op amps that don't suck.. several are characterized to drive 600 ohm within 100 mV of rails at 5V supply. I can tolerate a 10nV/rt hz ein.

1.1V/uSec slew rate may not seem fast, but I only need to swing a few volts p-p so still hundreds of kHz power bandwidth.  Right now I am leaning toward TI lmv722  $1.66 but there are a few in that ballpark with similar prices. I suspect that is a new competitive market for op amps.

JR

PS: Sorry about the veer... don't use a 5V CMOS op amp inside a tube amp... ::)
 
JohnRoberts said:
The 4562 is nice at higher rail voltage but the distortion plots for +/-2.5V start to rise well before 1 VAC. A larger concern is 10-12 mA per (package or per amplifier?), of course less current from 5V rail.  The OPA2134 was the really good op amp I mentioned but looks marginal drive capability for rail to rail use at 5V rail

John, for a 5V rail I recommend the AD8656, its a CMOS opamp for low voltages, it has 2.6 nV/rtHz, 28MHz bandwidth, 50uV of offset voltage, and 8.5 V/usec SR , +-220mA of output current and 3.7mA of current consumption Typ, seems to be rail to rail aswell, can go up to 4.991V and down 8mV with a 5V PSU.
 
scott2000 said:
Any opinions on the LM358?

The LM358 is not optimized for audio, however distortion can be reduced by adding an output pulldown resistor https://tangentsoft.net/audio/opamp-bias.html, the advantage of the 358 is that it can go down to practically the negative supply, SR is merely 0.55 V/uSec and Bandwidth is 1.2MHz, if that is not enough to convice you, the noise spec is 50nV/rtHz@1KHz, 350nV/rtHz at 10Hz  :eek:. It was (is?) the preferred choice of digital designers when they needed to operate an opamp without adding +/-V supplies.
 
user 37518 said:
is the TLV2172 you recommended also going to work without those?
Actually I don't know. The datasheet says yes but it looks like boilerplate language. I've never actually used TLV2172. I'm just pointing out that it's the active TI part that looks most like TL072. But in this particular case, AuxTex64 should just use TL072. I don't recommend substituting any other amp there.
 
user 37518 said:
John, for a 5V rail I recommend the AD8656, its a CMOS opamp for low voltages, it has 2.6 nV/rtHz, 28MHz bandwidth, 50uV of offset voltage, and 8.5 V/usec SR , +-220mA of output current and 3.7mA of current consumption Typ, seems to be rail to rail aswell, can go up to 4.991V and down 8mV with a 5V PSU.
The only thing I could find wrong with it was the price, and since I am building exactly one of these, I can live with $4 for a dual. 

Thanks it looks like a solid part for my application.

If I was going to sharp pencil this design I would steal a higher voltage PS from the amp modules and use common op amps at higher voltage, but this allows me to KISS and use with same 5V rail.

scott2000 said:
Any opinions on the LM358?
The LM358 (2904) is another old soldier that IMO wasn't a winner back in its day... IIRC it was designed for GP use in single supply designs (input range included ground (or -V).

I never used one.

JR
 
Back
Top