EEVblog 1752 - WTF Texas Instruments BIG NE5532 CHANGES

[aeg]

Fabs offering older processes are closing and TI is forced to move to a newer process? This would be the worst option. In this scenario Texas Instruments just happened to be the first and with time other manufacturers will follow.

The old and new 5532 are both made at TI's own fabs!

[Kleinstein]

As they are kind of ending the old process, other part should be effected too from changes. It looks like this also applies to the gold old uA741:
The new 2026, rev. H datasheet gives quite some changes:
no more offset adjust - though some pictures still show it, there are other oddities with the bias current
much lower supply current (0.13 mA compared to 1.7 mA typ)
much lower input bias , higher input resistance
near RR output swing
80 mA typ output current
+-18 V supply only   -- AFAIR they had a +-22 and +-18 V max version. At least for the LM741 they have.
--> not sure if this is still a BJT based part. Most specs are an improvement, but some could still break old designs.
     the list the TLV9301 as a similar part and from the specs - it could be a similar CMOS part.

On the positive side there is still the LM741 listed with more the original specs. The question may be for how long.

[chilternview]

The old and new 5532 are both made at TI's own fabs!

I believe they are different lines though. See Daves video where the PCN is shown, they were on SFAB process J11 150mm wafers, new on RFAB process TIB 300mm wafers.

[squadchannel]

As they are kind of ending the old process, other part should be effected too from changes. It looks like this also applies to the gold old uA741:
The new 2026, rev. H datasheet gives quite some changes:
no more offset adjust - though some pictures still show it, there are other oddities with the bias current
much lower supply current (0.13 mA compared to 1.7 mA typ)
much lower input bias , higher input resistance
near RR output swing
80 mA typ output current
+-18 V supply only   -- AFAIR they had a +-22 and +-18 V max version. At least for the LM741 they have.
--> not sure if this is still a BJT based part. Most specs are an improvement, but some could still break old designs.
     the list the TLV9301 as a similar part and from the specs - it could be a similar CMOS part.

On the positive side there is still the LM741 listed with more the original specs. The question may be for how long.

I'll post the PCN in another thread.

[The Chump]

This is some text from the private mail I received from Chris Featherstone on the TI e2e forum

For the transition to the new FAB the die was replaced with the RC4580. The listed Qual device in the PCN is the RC4580. For this reason you are seeing differences.
https://www.ti.com/lit/ds/symlink/rc4580.pdf?ts=1773855963647&ref_url=https%253A%252F%252Fwww.google.com%252F

For most specs on these replacements the die generally meets or exceeds the performance. There are cases such as this one that we are running into where there are typical specs (meaning not a guaranteed min or max spec) that are worse under certain conditions. The company is taking feedback over the last year on these transitions seriously.

[Alex Nikitin]

This is some text from the private mail I received from Chris Featherstone on the TI e2e forum

For the transition to the new FAB the die was replaced with the RC4580. The listed Qual device in the PCN is the RC4580. For this reason you are seeing differences.
https://www.ti.com/lit/ds/symlink/rc4580.pdf?ts=1773855963647&ref_url=https%253A%252F%252Fwww.google.com%252F

For most specs on these replacements the die generally meets or exceeds the performance. There are cases such as this one that we are running into where there are typical specs (meaning not a guaranteed min or max spec) that are worse under certain conditions. The company is taking feedback over the last year on these transitions seriously.

 

That confirms the TI management and engineering completely lost the plot a while ago. The only possible course of action left is to avoid TI products manufactured in 2023 or later altogether.

Cheers

Alex

[Agent24]

TI's management is giving big clown energy here alright.
Relabelling one chip die as another is the kind of shit I expect from shady Chinese vendors!

[Just Brew It!]

Before the NE5532 thing blew up and I found this thread, I was not aware of the TL07x change either. Worse noise performance on the new TL07x really sucks for anyone repairing old gear. And sure enough, all references to JFETs have been removed from the most recent data sheet.

[tooki]

The old and new 5532 are both made at TI's own fabs!

I believe they are different lines though. See Daves video where the PCN is shown, they were on SFAB process J11 150mm wafers, new on RFAB process TIB 300mm wafers.

Not just on different lines, SFAB and RFAB are two completely different fabs. SFAB is already decommissioned and is up for sale.

[santiall]

Before the NE5532 thing blew up and I found this thread, I was not aware of the TL07x change either. Worse noise performance on the new TL07x really sucks for anyone repairing old gear. And sure enough, all references to JFETs have been removed from the most recent data sheet.

well, look at the title: L07xx Low-Noise, FET-Input Operational Amplifiers
one may argue that a MOSFET is a FET but hey, it is super fishy to pull these tricks... at least just call it CMOS, not FET

be very careful with the TL07x as they changed them from JFET to CMOS technology and the noise spec is much worse now, 37nV/sqrtHz instead of the usual 15nV
At lest they changed the suffix to H but still very tricky.

They initially had the seprate making with the H for the CMOS version. However they now make it hard to tell if you get the 40 V max CMOS or 36 V max JFET version. With the destinction in the maximum specs the new parts are now all CMOS except for SO wide cases and the military version.
There was also a confusing PCN on the markings and versions. With the old fab changed they should only have a limited supply of the old JFET version.

They really need to add a chapter on how to tell the different chip versions apart.

that's right, what I mean is that it is 'easy' to specify in the BOM 'No H variant' and that pretty much eliminates Texas from the purchasing options. There are still plenty of TL07x made from other manufacturers with the expected/old specs.

Same for the 5532, just write the 'No Texas Instruments' and that should also remove any potential issues althought, I have to admit, I don't think I've ever used the 5532 over +/-15V although the slew rate can be a big problem. That doesn't hide the fact that degrading performance is a big no-no, at least call it TI15532 or whatever so one knows it is different.
I personally tend to stick to the 4580 or the 3307x when I need a low noise bipolar opamp for audio anyway.

[Just Brew It!]

Before the NE5532 thing blew up and I found this thread, I was not aware of the TL07x change either. Worse noise performance on the new TL07x really sucks for anyone repairing old gear. And sure enough, all references to JFETs have been removed from the most recent data sheet.

well, look at the title: L07xx Low-Noise, FET-Input Operational Amplifiers
one may argue that a MOSFET is a FET but hey, it is super fishy to pull these tricks... at least just call it CMOS, not FET

Well... a MOSFET is indeed a FET, and unless they're using complementary pairs it's not really CMOS. Calling it FET is "technically correct" regardless, but it is still really scummy that they changed the design in a way that degrades at least some of the specs, and still used the same part number.

I personally tend to stick to the 4580 or the 3307x when I need a low noise bipolar opamp for audio anyway.

You're probably OK then, since someone up-thread indicated that the "new" 5532 appears to be a re-badged 4580!

[santiall]

Before the NE5532 thing blew up and I found this thread, I was not aware of the TL07x change either. Worse noise performance on the new TL07x really sucks for anyone repairing old gear. And sure enough, all references to JFETs have been removed from the most recent data sheet.

well, look at the title: L07xx Low-Noise, FET-Input Operational Amplifiers
one may argue that a MOSFET is a FET but hey, it is super fishy to pull these tricks... at least just call it CMOS, not FET

Well... a MOSFET is indeed a FET, and unless they're using complementary pairs it's not really CMOS. Calling it FET is "technically correct" regardless, but it is still really scummy that they changed the design in a way that degrades at least some of the specs, and still used the same part number.

I personally tend to stick to the 4580 or the 3307x when I need a low noise bipolar opamp for audio anyway.

You're probably OK then, since someone up-thread indicated that the "new" 5532 appears to be a re-badged 4580!

sure, we are on the same page regarding technicalities about the name although a JFET isn't the same as a CMOS. My point is that the TL07x were lanched and marketed back in the day as a low noise JFET input opamps and other opamps were marketed as CMOS but now they slide the thing in with worse specs mainly regarding noise, which was one of the main features...

I also have the impression that they might be consoilidating manufacturing processes, using the same wafer for different products. At the end the TL07x and the 5532 are old-junk, aren't they? , they'll tell us that there are better/more modern options anyway

[golden_labels]

Time to get a die photo of the new TI NE5532 and compare to a true one⁽¹⁾, and to Ti RC4580?

⁽¹⁾ Two examples from Mikhail of ZeptoBars: Ti NE5532 - real vs fake (2017), Fake audio opamp - Ti NE5532 remarked as AD OP275 (2019).

[Kleinstein]

The RC4580 is also noted in the PCN. So there could also be an old and new RC4580 version from Ti.  At least in this case the changes seem to be small.
To somewhat replace the NE5532 they would need a slightly lower noise - a change nobody would really mind.

[iMo]

The old and new 5532 are both made at TI's own fabs!

I believe they are different lines though. See Daves video where the PCN is shown, they were on SFAB process J11 150mm wafers, new on RFAB process TIB 300mm wafers.

The RFAB operates since 2009 afaik, so there will be plethora of older analog parts migrated to the new process, imho..

[Bzzz]

Well, my expectations for TI were low since this dick move back in 2023.
Discontinued the ULN2803A - introduced the ULN2803C.
Great, what has changed? They have added two NC pins at the end.

[Agent24]

Well, my expectations for TI were low since this dick move back in 2023.
Discontinued the ULN2803A - introduced the ULN2803C.
Great, what has changed? They have added two NC pins at the end.

What!?!

[Bzzz]

Exactly my reaction.

https://www.farnell.com/datasheets/48660.pdf <-- old
https://www.farnell.com/datasheets/3775927.pdf <-- PCN
https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/1134115/faq-uln2803c-uln2803adwr-uln2803cdwr-creating-a-common-pcb-footprint-for-18-pin-and-20-pin-8ch-darlington-transistor-arrays-devices <-- TI dev FAQ
https://www.ti.com/lit/ds/symlink/uln2803c.pdf <-- new

All of our layouts can fit the 20p part physically, but on one of them, the additional bit sticks out. Go ahead and tell the assembly house and later on the customer that this is intentional...

[tszaboo]

Well, if I was them, I would've sold the old factory as-is. Just spin off the old factory, sell it together with the machines, make a new company. Call it, IDK, Burr Brown or something.
And call your new opamp NE5532T or whatever.

[squadchannel]

Let's trim it with a dremel.

[madires]

Well, my expectations for TI were low since this dick move back in 2023.
Discontinued the ULN2803A - introduced the ULN2803C.
Great, what has changed? They have added two NC pins at the end.

You can't make that up. Anyhow, I guess it's the result of a cost optimization.

[iMo]

Well, if I was them, I would've sold the old factory as-is. Just spin off the old factory, sell it together with the machines, make a new company. Call it, IDK, Burr Brown or something.
And call your new opamp NE5532T or whatever.

The old TI's SFAB factory won't be retired after all. Rumor has it it's getting loaded onto Starship Flight 19 and sent to space as a proof of concept.

[Agent24]

Well, my expectations for TI were low since this dick move back in 2023.
Discontinued the ULN2803A - introduced the ULN2803C.
Great, what has changed? They have added two NC pins at the end.

You can't make that up. Anyhow, I guess it's the result of a cost optimization.

Feels more like the result of junior engineers using ChatGPT actually, because no sane person would come up with such a pointlessly batshit idea.

[Agent24]

https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/1134115/faq-uln2803c-uln2803adwr-uln2803cdwr-creating-a-common-pcb-footprint-for-18-pin-and-20-pin-8ch-darlington-transistor-arrays-devices <-- TI dev FAQ

"We made the IC have 20 pins now so it's compatible with similar 18-pin devices, you just ignore the bottom 2 pins we just added and align the part with Pin 1 as normal"

Am I missing the point here because this really sounds like ChatGPT delulu engineering.
You can leave the footprint at 18 pins and it's still compatible with other similar 18 pin parts, surely!?!?
What the actual...?

[Kleinstein]

The SO20 case is a lot more common than SO18. Though more material, it could be the slightly cheaper option and for a new design the extra pins improve heat sinking. At least they gave it a new name as the change in the dimensions is obvious and a bit more difficult to fool the customers on this. It should be just a new case option.